REPORT of INVESTIGATION
into the
MINE EXPLOSION at the

UPPER BIG BRANCH MINE
APRIL 5, 2010
BOONE / RALEIGH CO., WEST VIRGINIA

WEST VIRGINIA OFFICE of MINERS’ HEALTH,
SAFETY & TRAINING
FEBRUARY 23, 2012
C.A. PHILLIPS, DIRECTOR

 CONTENTS

1

Executive Summary

2

General Description and Setting

3

Operations: Pre-explosion

4

The Explosion

5

The Rescue and Recovery

6

Administrative Inquiry

7

Investigation into The Cause of the Explosion

8

Source of Fuel for the Explosion

9

Conclusions and Recommendations

10

List of Appendices

 1

Executive Summary

On April 5, 2010, at approximately 3:02 p.m., an explosion ripped through the underground
workings at Performance Coal Company, UBBMC, Montcoal Eagle Mine (UBB), taking the
lives of 29 miners and injuring two others.

The explosion began after gas, mostly methane, was ignited by frictional impact as the shearer
was cutting sandstone roof or by the rock colliding with steel supports or other rock while falling
from the sandstone roof behind the longwall shields. It is deemed more likely that the longwall
mining machine (shearer) ignited the methane.

The methane gas apparently was liberated from the mine floor behind the roof-supporting
longwall shields, and, as this gas flowed to the return behind the shields, the airflow then became
restricted by a recent roof fall across the #7 entry of the #21 Tailgate. This may have allowed
combustible gas to accumulate near the location where the shearer was operating. This roof fall
would not allow the air current to move directly toward the return at the tailgate corner of the
gob, causing it to flow outby to an open crosscut. The shearer shut down just a few feet short of
its normal stop position as it mined into the tailgate entry. Approximately 1½ minutes later the
methane explosion occurred in the gob behind the longwall shields, spreading quickly to the #21
Tailgate and from there propagated through an extensive area of the mine.

This accumulation of methane was not detected by the required mine examinations or by the
required machine-mounted methane monitors. Because the “T-split” is a critical location where

1

EXECUTIVE SUMMARY

 ventilation can become blocked and gases can quickly accumulate, additional measures to
support the roof at this location are prudent and recommended. More effective monitoring of
gases at this location is also advised.

The methane explosion quickly transitioned into a coal dust explosion, which severely damaged
ventilation controls, conveyor belts, water lines, electrical systems and numerous items in its path
until the fuel was consumed and the explosion extinguished itself outby the track switch at the
beginning of North Glory Mains.

A thorough investigation was conducted and the findings documented in this report to the
Director of the West Virginia Office of Miners’ Health, Safety & Training (WVOMHS&T).
Contained in the body of this report is a summary of our findings. Details related to the data
developed and the interpretation thereof can be found in the report appendices.

This report is the product of a 20-month investigation which included a thorough underground
examination, including the examination of electrical components in the mine, evaluation of the
mine geology, comprehensive mapping as part of the joint investigation and supplemental
mapping by WVOMHS&T. Interviews were conducted, evidence was collected and evaluated
and thousands of documents were examined.

There were certain areas of the mine which were inaccessible for safety reasons, but the
remaining areas of the mine comprising a cumulative distance of mine passageways of over 50
miles in length were examined thoroughly and documented extensively in our investigation. The
accessible entries and crosscuts were each traversed and examined multiple times in the course of
our investigation.

Part of our investigation involved examining the sources of fine coal dust, both float dust from
mining sections and conveyor belts, as well as sloughed material from friable coal ribs. We find
that fine accumulations that can easily be dispersed into the air by the pressure wave ahead of a
propagating flame are of primary concern and deserve increased attention.

2

EXECUTIVE SUMMARY

 The amount of rock dust being maintained on mine surfaces at the time of the explosion was
insufficient to stop a coal dust explosion. The region where the dust explosion started does not
appear to have had rock dust periodically applied over the fine coal dust. Periodic applications of
rock dust over accumulating fine coal dust are necessary to render such dust harmless.

The greatest defense against the hazards of fine coal dust is the proper application of rock dust.
The way we currently sample and apply rock dust needs to be modernized, using the latest
research knowledge and most appropriate sampling and analyzing methods available. Other
strategies to arrest a dust explosion include explosion mitigation barriers, which have important
applications in certain conditions. Investigators found evidence that a water barrier just west of
#22 Cross-over stopped the explosion from propagating further inby in the #21 Headgate entries.
Further research is needed to demonstrate the practical application of water barriers, rock rubble
barriers and other explosion-mitigating strategies as supplemental protection with generalized
rock dusting to prevent explosion propagations in the future. The data combined in this report
will serve as a resource for future research.

3

EXECUTIVE SUMMARY

 2

General Description and Setting

2.1

Location

2.2

Geology and Mining Conditions

2.3

Method of Mining

2.4

Ventilation and Methane Drainage

2.5

Communication and Tracking System

2.6

Organizational Structure

 2
2.1

General Description and Setting

Location

Figure 1. The Upper Big Branch Mine is located in Boone and Raleigh counties in West Virginia.

Performance Coal Company’s UBBMC Montcoal Eagle mine (UBB) is located in Montcoal,
just off West Virginia Route 3 in Raleigh County. UBB is 46 miles from the capital city of
Charleston, 34 miles from Beckley and 45 miles from Oak Hill. The nearest emergency service
agency is located in Whitesville, West Virginia, six miles from UBB. The North Portal of UBB
is located at latitude N 37º 55’ 2.04” and longitude W 81º 33’ 8.10”.

1

SECTION 2

 Figure 2. Map showing location of portals and fans. See inset map, below for additional information.

Figure 3. Map showing the names of selected locations and headings referred to in this report.

2

SECTION 2

 2.2

Geology and Mining Conditions

The Eagle seam is a premium metallurgical “mid-vol” coal of Middle Carboniferous age and is
stratigraphically a part of the Kanawha Formation of the Pottsville Group (see Appendix 2.2-1).
The roof rock at UBB is typically sandstone, but occasionally is shale. The sandstone is gray,
medium fine grained and composed mostly of quartz grains (67%). Near the longwall face the
first few inches of the immediate roof are sometimes a hard, brown sandstone which contains
the sulfide mineral pyrite and/or marcasite,1 but in small amounts (less than 3.5%; see
Appendix 2.2-2).

Massive SS

0 - 36"

Gray shale

12 - 20"
1- 40"
15 - 30"
2 - 10"

10 - 15 ft.

10 - 20"
40 - 80 ft.

Approx. 190 ft to
POWELLTON SEAM

COAL; Bright, soft,
fusain partings in middle

TOP BENCH

Carbonaceous shale

Parting

COAL; Dull, hard

BOTTOM BENCH

Boney coal

EAGLE SEAM

30 + ft.

Sandy shale; rooted

COAL

LITTLE EAGLE

Shale under coal, then
mostly sandstone

BETSIE SHALE

Figure 4. Generalized stratigraphic section showing the seam mined at UBB (Eagle seam), its
immediate roof rock (typically sandstone, occasionally shale), and the location of the underlying leader
seam (Little Eagle seam), and Betsie Shale.

1

Pyrite and marcasite are chemically the same (FeS), but have different crystalline structures.

3

SECTION 2

 Marcasite was used by Paleolithic and some Native Americans for striking fires,2 but it is
unknown if the amounts present in the sandstone give it enhanced sparking characteristics. The
brown sandstone contains half the quartz and twice the amount of albite (a feldspar mineral) as
does the gray sandstone.
Seams immediately above the Eagle seam
The first mineable seam above the Eagle seam is the Powellton seam, which lies approximately
180 feet overhead and is actively being mined (see Appendix 2.2-1). Overall, the mine roof at
UBB is massive, medium grained sandstone that is generally competent. Local areas of bad top
occur where roof rock changes from sandstone and are also due to localized accumulations of
plant fossils in the sandstone, which create bedding separations that produce occasional dropouts.
The immediate 1 to 2 feet of the mine floor is typically sandy shale, weakened somewhat by
fossil root structures. Typically the floor is softer than the roof, so when additional mining
height is desired it is the floor rock that is taken. Floor rock had been mined at the tailgate
during development, and was also being mined by the shearer. Below the immediate floor is
sandy shale and sandstone.
Strata immediately below the Eagle seam
Approximately 10 to 15 feet below the Eagle seam is the Little Eagle seam, which is typically
10 to 20 inches thick and free of rock partings. Below the Little Eagle is the Betsie Shale, which
has a thickness of 100 to 250 feet; core logs show that the top half is not shale but sandstone.
The Betsie could be a source and reservoir for the accumulation of local pockets of natural gas,
but additional research is needed (see Appendix 2.2-1).

2

New York State Museum Bulletin 511, 2009, Chapter 8, Algonquian and Iroquois Uses of Plants and Other
Materials to Make Fire

4

SECTION 2

 Characteristics of the Eagle seam
The Eagle seam has a Top Bench of 12 to 20 inches of coal that is vitreous, fractured and
friable. The center 4 to 6 inches of this bench contain approximately 1.5 inches of distinctive
“mother coal” 3 that occurs sometimes as a single layer, but more typically as 3 to 6 thin, silvery
and sooty partings. This “fusain” crumbles easily to fine dust, much of which readily passes
through a 200 mesh sieve, and, despite its low volatile content, it is comparable in explosibility
to Pittsburgh seam float dust (see Appendix 8.2-1).
A shale parting separates the (Eagle) Top Bench from the Bottom Bench. Its thickness varies
between 1 and 40 inches, sometimes erratically over distances of a few dozen feet.
The (Eagle) Bottom Bench is a hard and splinty coal and varies between 15 to 30 inches in
thickness, with an additional 2 to 10 inches of boney coal (very hard) at the very bottom.
The (Eagle) Top Bench is the softest material in the seam section so that when the pillars take
weight it fractures readily and spalls out, leaving cavities and ledges. These ledges can catch
and accumulate dust from air currents and from progressive spalling. The shale parting that
separates the two benches is soft when its thickness is 6 inches or less and is firm when its
thickness exceeds about 16 inches. The (Eagle) Bottom Bench is a very hard coal and generally
stands firm, as does the floor and roof, so that the weakest strata generally is the Top Bench of
the Eagle seam; however, “floor heave” occurs in a variety of places. The depth of cover is
generally greater than 800 feet throughout the study area. Floor heave is the term applied to
fractures in the floor that have opened up and exhibit vertical displacements of usually 1 foot or
less but occasionally as much as 3 feet. They tend to be most prevalent where mine entries are
aligned N-S or E-W.

3

Fusinite and semi-fusinite also, commonly referred to as fusain.

5

SECTION 2

 Local fractures and gas accumulations
Structurally, the dominant roof fractures are natural extension fractures trending along
approximately N 25º W. This is similar to the mine floor, which overall has a dominant fracture
trend of approximately N 20º W. The coal face cleat is equivalent to the rock extension
fractures but is on a different orientation of approximately N 80º E. Fracture systems of joints
and cleats are likely the dominant way that gases from below migrate upward.
The floor strata (which include the Little Eagle underclay, the Little Eagle seam and the
interburden between the Little Eagle and the Eagle seam) may serve as a cap, preventing the
natural upward migration of gases from the Betsie Shale or lower source beds. Equilibrium
between gas pressure and confinement pressures are sometimes disrupted by mining, allowing
localized gas to enter the mining spaces via the natural fracture system.
For more information about the geology, see Appendix 2.2-1.

2.3

Method of Mining

The following types of equipment and mining methods were used at UBB in the production of
coal and other support activities at the time of the explosion on April 5, 2010.

The development sections worked two production shifts a day, mining coal at least five days a
week, and sometimes six days a week, with a maintenance shift on the midnight or “owl” shift.
The longwall normally produced coal seven days a week with three crews rotating the
production schedule. The longwall maintenance and support crews on the midnight shift did not
rotate. Starting times varied for production and non-production crews.

6

SECTION 2

 The mine was ventilated with three fans. Two blowing fans, one located at the North Portal and
the other located at the South Portal. The third fan, located west of the #21 Headgate Longwall
at Bandytown was an exhausting fan.

The production units on the advancing room and pillar sections consisted of Joy 12CM and
14CM remote control continuous mining machines. The roof was supported with Fletcher RRII
and CDRII-13 dual boom roof bolting machines. The coal was transported to the conveyor belts
with Joy 10SC shuttle cars. Fairchild and Long Airdox battery operated scoops were used to
assist in the cleanup of the working faces, to move supplies on the sections and support other
activities throughout the mine. A few battery powered A. L. Lee forklifts also were used to
assist in the movement and transportation of mining supplies in outby areas of the mine.

The two longwall development sections, #22 Tailgate and #22 Headgate, utilized two
continuous miners, three shuttle cars and two dual-head roof bolting machines on three entry
sections. These two longwall development sections were using sweep ventilation, and mined the
section using the “walking style” method of mining. The walking style of mining allowed the
extra face equipment to be repositioned for mining while the other equipment was producing
coal. These two sections and the longwall section transported the coal by conveyor belts out the
Ellis Portal.

The other room and pillar section located in the north side of the mine was the Barrier section,
located between the partially mined #16 and #17 Headgate Longwall panels. This seven entry
section also utilized two continuous miners, two dual head roof bolting machines and three
shuttle cars for production purposes. Air in this section was swept from right to left. The coal
from the Barrier section was transported by belt out the East Mains or Silo Portal.

The longwall also located in the north side of the mine, was mining the #21 Headgate panel in an
up dip direction, retreating west to east, at the time of the explosion. The longwall had retreated

7

SECTION 2

 approximately 5,400 feet as of April 5, 2010. The longwall face setup consisted of a Joy 7LS
shearer, Joy Shields and stage loader unit. The longwall air was intaking down the face and
returning out towards the Bandytown fan located west of the longwall face.

The Portal section, located in the south side of the mine, also used two continuous mining
machines, two dual head roof bolting machines and three shuttle cars in the production of coal.
This section, however, was the only one that ventilated with “split or fishtail ventilation.” This
section was not working the day of the explosion and had been “Idle” in the fireboss books since
March 29, 2010. Plans were underway to relocate this section.

The #22 Headgate section working faces were the deepest workings, measuring approximately
4.25 miles from the Ellis Portal; the #21 Longwall face was approximately 3.5 miles from the
Ellis Portal.

The mine utilized rail haulage from the North, South and Ellis portals to transport workers and
supplies into the mine using Brookville battery powered motors and mantrips. Conveyor belts
were used to transport the coal from underground. UBB utilized a complex belt system to carry
the coal from the Ellis and East Mains or Silo portals. The conveyor belt that exited the Ellis
Portal went across West Virginia Route 3 and back underground through other mines to the
Marfork Preparation Plant, located southwest of UBB. The conveyor belt that exited the East
Mains or Silo Portal also traveled across West Virginia Route 3 and tied in with the same belt
system that ends at the Marfork Preparation Plant.

8

SECTION 2

 2.4

Ventilation and Methane Drainage

Mine ventilation is the most important of the defenses to prevent mine explosions and respirable
dust diseases. Three fans were installed on the surface; two were installed at drift openings
blowing air into the mine and the third was an exhaust fan. Most of the air exhausting from the
mine was via a 16-foot diameter airshaft near Bandytown. This fan was assisted by blowing fans
at the North and South portals (see Figure 5). These three fans moved a large quantity of air,
about 1,013,900 cubic feet per minute (cfm), of air was reported to be returning to the surface
after having ventilated the underground areas. Ventilation is essential to dilute and remove
noxious gases, methane and respirable dust.

Figure 5. General layout of the ventilation at UBB at the time of the April 5, 2010 explosion.

9

SECTION 2

 The air currents are directed as needed to various locations underground, where needed, by
concrete block walls (stoppings) strategically placed in mine openings. Three separate air
currents are used: intake, belt air and return. Air crossings (overcasts) were used where the
various air currents, such as intakes and returns, crossed as they were divided throughout the
mine. A system of airlock doors were used to separate one type ventilation current, such as a
current dedicated to belt air, and another type, such as the intake.

The Federal Mine Safety and Health Act contains three categories of methane liberation: more
than 200,000, over 500,000 and over 1,000,000 cubic feet of methane liberated in a 24-hour
period. The UBB mine liberated about 1,067,510 cubic feet of methane per day according to the
last available air measurements and air analysis. About 981,000 cubic feet of the methane was
reportedly exhausting from the Bandytown fan. Air analysis found 0.182% methane in the air
current and about 0.11% carbon dioxide (normal air contains about 0.03% carbon dioxide).
Significant amounts of the methane came from the rocks below the Eagle coal seam, as stress
fields changed during mining, and probably some came from the roof as it caved. Outbursts of
gas occurred from breaks in the mine floor near longwall shields on February 18, 2004, and July
3, 2003.

A large portion of the previous mining had been sealed at the time of the explosion (see Figure
5). The active portions of the mine were north and east of that sealed area. Carbon dioxide,
methane and air with low oxygen are gases normally found in sealed gobs. Even though the seals
are competent, some gases can escape around their perimeters during periods when the barometer
trends downward. For that reason, the seals were ventilated by air directed to the return and
exhausted at the North and Silo portals. Most of the coal production was from the region
ventilated by the Bandytown fan north of the large sealed area. This is the area affected by the
explosion, where one longwall section and two continuous miner sections were operating.

10

SECTION 2

 2.5

Communication and Tracking System

The communication and tracking system approved and installed at UBB consisted of a Minecom
UHF leaky feeder-type communication system that utilized wireless handheld Motorola radios
and Pyott-Boone, Model #1981 tracking boss, tag readers. The tag readers are located at
specific locations underground. The RF ID Tags, Model #1980, were worn by individual miners
and used to track their location while underground. The tracking system was tied in with the
Pyott-Boone carbon monoxide mine-wide monitoring system (MineBoss), which was already in
use at the mine.

An Emergency Communication and Tracking Plan for UBB was approved on January 22, 2008
(see Appendix 2.5-1). Violation #26611 was issued on August 4, 2009, by the WVOMHS&T
for not being in compliance with the approved plan. This violation was extended at least five
times through March 1, 2010, as work on this system progressed (see Appendix 2.5-2).

At the time of the explosion, the tracking system was still not fully operational throughout the
entire mine. Work was being done to complete the tracking system and also to eliminate
electrical problems (blown fuses) in the system amplifiers. The Tag Reader Tracking System
was operational from the surface up to crosscut 102 North Mains (near the mother drive). All
other areas of this mine not covered by the operational tag readers were required to be tracked by
manual tracking procedures.

The tag reader (station ID 3.5), located at crosscut 76 on the North Mains Track, identified
several people that were in the #22 Tailgate Section mantrip at approximately 3:01.53 p.m.,
minus eight seconds for data transfer, or 3:01.45 p.m. on April 5, 2010.

11

SECTION 2

 At approximately 3:02.18 p.m., the Pyott-Boone system computer, located on the surface,
recorded a communication loss from a CO monitor located at crosscut 99 on 6 North Belt as the
explosion severed a power cable near that location (see Section 4.2). Due to the programming
and data transfer speed of the system, the event time may be as much as 108 seconds prior to the
recorded time of the alarm, or 3:00.30 p.m.4 This system was rendered inoperative by the
explosion on April 5, 2010 (see Appendix 2.5-4).

2.6

Organizational Structure

Performance Coal Company is a West Virginia corporation and a 100% subsidiary of Elk Run
Coal Company, which is also a West Virginia corporation. Elk Run Coal Company is a 100%
subsidiary of A. T. Massey Coal Company, Inc. which is a Virginia corporation. A. T. Massey
Coal Company, Inc. is a 100% subsidiary of the parent Massey Energy Company, a Delaware
corporation.

On April 5, 2010, the corporate officers of Performance Coal Company were Mark Clemens,
Director; Chris Blanchard, Director/President; Jamie Ferguson, Vice-President; Richard
Grinnan, Secretary; Shane Harvey, Assistant Secretary; and Phillip Nichols, Treasurer.

The management structure at UBB consisted of Wayne Persinger as general manager. Everett
Hager was superintendent and Terry Moore was the mine foreman on the north side of the mine.
Gary May was the superintendent and Rick Foster was the mine foreman of the south side of the
mine. Jack Roles was the longwall coordinator.
4

Time after time drift correction was conducted by WVOMHS&T (see Appendix 2.5-3).

12

SECTION 2

 3

Operations: Pre-explosion

3.1

History and Previous Mining

3.2

Incidents of Prior Gas Inundations

 3
3.1

Operations: Pre-explosion

History and Previous Mining

UBB is located near the community of Montcoal, in the Marsh Fork District of Raleigh County.
“Advance Approval” from the West Virginia Division of Environmental Protection, later known
as Department of Environmental Protection (WVDEP), was transferred from Peabody Coal
Company to Performance Coal Company on October 7, 1994. The WVOMHS&T issued the
Health and Safety permits to Performance Coal Company, Upper Big Branch Mining Company,
Montcoal Eagle Mine (UBB) on November 2, 1994. UBB was assigned permit number U-304292. There was no underground mining activity on the permit prior to Performance Coal
Company acquiring this property.

Production began and within approximately 18 months additional openings had been developed
for ventilation, transportation of workers and coal. This was completed with the use of remote
control continuous miners, shuttle cars and double head roof bolting machines that utilized the
room and pillar method.

Mining of the West Mains and Hazy Mains was developed to set up longwall panels and a
bleeder system to ventilate the 10 projected longwall panels in the reserves on the south side of
the mine. The longwall panels were developed to allow the retreating face to mine in a west to

1

SECTION 3

 east direction. The first longwall panel was completed by March 1997. Additional openings for
ventilation and transportation of coal and workers were developed as mining advanced. The
longwall completed mining 10 panels in the south area of the mine by March 2000. Later that
month, the longwall was moved to the north reserves where nine panels were mined. Three of
these panels were not totally mined on initial setup. Those three partial panels were mined in
2005 and 2006.

Longwall mining at UBB ceased in July 2006 when the short panel on #16 Headgate was
completed. Since there were no other panels developed at UBB, the longwall was moved to the
Castle Mine (Logans Fork) of Elk Run Coal Company while additional panels in the north
reserves of UBB were being developed. The longwall returned to UBB prematurely in the
summer of 2009 due to difficult mining conditions at the Castle Mine. This unexpected
development caused some issues with the setup area for the returning longwall. Ventilation and
water drainage issues had not been completely resolved prior to the longwall being put back in
production at UBB in September 2009.

On April 5, 2010, UBB was operating four production sections in the Eagle seam. The Barrier
section was located between the old #16 Headgate and #17 Headgate panels. The longwall was
producing on the #21 panel. The #22 Headgate and #22 Tailgate sections were located north of
the longwall and were developing a future longwall panel. This longwall panel was behind
schedule and would not have been ready when the current panel was completed. As a result,
plans were to develop a short panel at the Ellis/LBB Construction site.

Three coal mines are adjacent to UBB in the Eagle seam. On the southwest side is the Peabody
Coal Company, Harris No.1 Mine. The abandoned Eagle Energy Inc., No.1 Mine is located on
the west side. The third mine, Ellis Eagle Mine of Marfork Coal Company, is located on the east
side across West Virginia Route 3. Longwall mining was conducted at these mines.

2

SECTION 3

 Production History

The UBB mine produced approximately 13,000 tons of coal in December, 1994. When the
longwall went into service in April 1996 coal production increased to just over 3,080,000 tons for
the year. This began a period of 10 consecutive years (1996 to 2005) that production exceeded
two million tons annually.

The highest production year was 1998 when approximately 5,698,800 tons were produced. From
1994 to 2010 the mine produced approximately 41,400,000 tons of coal.

Figure 1. Annual Production records from Upper Big Branch Mine, 1994 to 2010.

After the longwall was moved to the Castle Mine, production dropped to 576,672 tons in 2007
and approximately 365,000 tons in 2008. The longwall returned to UBB in 2009 and began
producing in September of that year.

3

SECTION 3

 Mining History of the Longwall

When operation of the south portion of UBB began in April 1996 panels were mined until those
reserves were depleted. The only reported event dealing with an ignition or gas inundation on a
longwall face while mining these reserves occurred on January 4, 1997.

The North Reserves were developed and longwall production began in March 2000. Prior to the
April 5, 2010 explosion, there were at least two occasions of large inundations of gas while
mining the North Reserves. There were no catastrophic repercussions with either incident. One
incident occurred in 2003 and another in 2004.
.
When the longwall section was moved to the overlying castle mine in July 2006 coal was
transported through the Castle Mine, down the Glory Hole and then by conveyor belts to the
surface of UBB. The Glory Hole was developed from UBB into the Castle Mine which is
approximately 180 feet above. Bad mining conditions at the Castle Mine were the catalyst to
bring the longwall back to UBB before originally planned.
UBB continued mining operations with continuous miners in the longwall’s absence. The #21
Headgate Longwall panel was developed at UBB and the longwall was brought back from the
Castle Mine. Longwall production at UBB resumed in September 2009. The #21 Longwall
panel was developed approximately 6,700 feet in length with a longwall face approximately
1,000 feet wide. This panel was being mined west to east, with a rise in elevation of
approximately 85 feet from the beginning of production until the time of the accident. The depth
of cover for the panel being mined was as much as 1,275 feet, and was approximately 1,075 feet
at the current face location. Mining had progressed approximately 5,400 feet at the time of the
explosion. The Bandytown exhaust fan was located approximately 11,600 feet west of the #21

4

SECTION 3

 Longwall face at the time of the explosion. A blowing fan also was utilized at the North Portal to
assist in mine ventilation. This panel was the first one to be ventilated by the Bandytown exhaust
fan. This fan was installed to ventilate the longwall gob for this and following panels to be mined
in this area, and to assist in ventilating the two working sections north of the active longwall
section.

Longwall Equipment and Systems

There were 176 Joy twin leg shields that made up the face support. The longwall face was mined
by a double drum shearer utilizing bi-directional mining as a regular part of the mining sequence.
Coal was transported along the face onto the stage loader by a steel conveyor chain. From there,
conveyor belts transported the coal to the surface by way of the Ellis Portal.

The longwall face methane monitor sensor was located at the tail of the longwall under the tail
drive motor covers, and the readout for the face monitor was at the headgate control box area.
The shearer also had a methane monitor and constant readout on the shearer.

Shield hydraulics was supported by Kamat Pumps with fresh water for the emulsion brought into
the mine from two deep freshwater wells. Water for the longwall shearer and other sections of
the mine was pumped from the Coal River by way of 150 horsepower pumps, and then stored in
large tanks located on the surface above the Silo or East Mains Portal. This water flows
underground from the tanks and, after being filtered, is transported underground by means of
PVC piping throughout the mine and to the longwall face. Pressure pumps are installed at
strategic locations to obtain the water pressures needed to operate the longwall shearer,
continuous miners and other water activated systems.
Pressurized water was provided to the shearer by a Sunflo Pump located at the “Mule Train” near
the end of the longwall section track.

5

SECTION 3

 3.2

Incidents of Prior Gas Inundations

Summary of Incidents
Between 1997 and 2004, there were at least three known gas events associated with longwall
mining at UBB. The first event led to a series of methane ignitions. The last two were gas
inundations, but no ignitions were involved. The event which occurred January 4, 1997 involved
one or more gas ignitions. Testimony from eyewitnesses indicates two or more ignitions
occurred at approximately 10:20 a.m., beginning in the gob subsequently entering the #2 West
Longwall face.

MAP 1. Location of the April 5, 2010 explosion, and prior gas events in 1997, 2003 and 2004.

6

SECTION 3

 January 4, 19971 Gas Ignition
The nine-man crew began production on Saturday morning, January 4, 1997, from the tailgate of
the longwall where the shearer was left the previous shift. Air measurements taken by the
foreman showed that airflows on the working face were in excess of the amounts required.
Workers stated that face ventilation seemed good, and face mining conditions seemed better than
normal. However, the mine roof was breaking higher up more than usual, and the mine roof fell
several times behind the longwall shields that morning.

The crew completed the fourth mining pass of the day by cutting out on the tail of the section at
about 10:20 a.m. As the operator was "shearing down” with the tail drum of the shearer, the
mine roof fell behind the shields and an ignition occurred in the gob area where the roof had just
fallen.

The shield operator had just advanced shields 175 and 176 and he was standing at shield 174
facing the gob and was first to see the ignition. He saw a red glow in the gob that was becoming
brighter. He pointed toward the glow, and then started running toward the longwall headgate.
He felt heat on his legs and ran 400 to 500 feet to about mid-face before getting out of smoke that
resulted from the ignition. Both shearer operators also saw the ignition, and flame singed hair on
the tail drum operator’s neck and arm. The head drum operator next saw the ignition come from
behind the shields and then continue up the face line from the longwall tailgate. These two
shearer operators also ran toward the headgate and said “something blew up.”

Two workers were changing batteries in the speaker phone near shield 150 when the first ignition
occurred. One felt heat from the ignition, but did not see any flash or flame. He thought the
shearer power cable had exploded so he called the headgate operator on the speaker phone and
had him de-energize the face power. The foreman and another worker went to the tail to check
1

See Appendix 3.2-1 (Report of a Methane Gas Ignition, January 4, 1997) for additional information.

7

SECTION 3

 the area after making sure everyone was headed out toward the head. The foreman stated he also
observed the ignition.

Both of these men detected 0.6% methane at shield 174, and enough carbon monoxide was
present to activate the alarm of the foreman’s LTX 310 detector. The detector was set to alarm at
50 parts per million (ppm) carbon monoxide. They detected a smell like “old works,” but no fire
was present. These two men witnessed a second ignition, which appeared like a “yellowish”
flash, while they were examining near shield 174. A third event occurred as they were walking
past shield 36 toward the longwall headgate. They described this final event as “bucking of the
air.”

The crew was directed to assure proper ventilation in the headgate entries, and a ventilation
curtain was also installed in the first four shields on the face to direct all available air down the
longwall face. The foreman started his exam back down the face but stopped when he got to
shield 92 where he experienced heat. The men then evacuated the mine.

The immediate roof in the tailgate return aircourse adjacent to shield 176 (the last shield) was
unstable and was blocked by a roof fall about 50 feet outby the tailgate of the section near survey
station 1266 (see Appendix 3.2-1 for map). The return air course was being carried as "dangered
off" in the fireboss reports for the last several days because of "bad top." It was also determined
during the investigation that the back-up curtains in the headgate entries were not installed and
being used at the time of the ignition.

The WVOMHS&T investigation determined that adequate airflow was not provided in the
longwall section bleeder system sufficient to prevent a dangerous accumulation of methane gas
in the gob area. It is believed that the gas was ignited by sparks created when the sandstone mine
roof struck the longwall shields after an unusually high roof fall.

8

SECTION 3

 Examination of the #1 fan chart shows a number of air pressure events occurred between 10 a.m.
and 11:30 a.m. on January 4, 1997. Samples taken by MSHA inspectors on January 4, 1997,
showed large amounts of coke at shields 172, 175 and 176, the tail motor and the shearer area.
Recommendations were given to address this ignition as part of the WVOMHS&T report
referenced.

July 3, 2003 Gas Inundation
The July 3, 2003, gas inundation was never reported to the WVOMHS&T. The information was
obtained through testimony given during the investigation of the April 5, 2010, explosion and
from information available from the MSHA Mine Data Retrieval System Website.

This 2003 event occurred on the #16 Longwall panel near shield 95 in the north reserves of the
mine. There were no injuries reported, and ventilation changes were made to help ventilate the
bleeder located in the mine floor in the gob behind the shields until the gas inundation subsided
and mining operations resumed.

February 18, 2004 Gas Inundation
A reported gas inundation occurred on February 18, 2004, on the #17 Longwall panel. Workers
heard a loud bumping or big thump at approximately 11:40 a.m. Wednesday morning followed
by sounds of gas pressure being released in the gob. The section was not producing coal at this
time due to a maintenance issue with the longwall shearer cable junction box.

Ventilating curtains were placed from the headgate to shield 87 along the jack line to assist in
diluting the floor bleeder gases to a manageable level. This floor crack was most prominent in
the mine floor around shields 106 and 107. The gas readings exceeded 5% methane at several
locations and at various times during this event. By 3:20 p.m. Friday, February 20, 2004, the gas
inundation had subsided and the control order regarding this incident was lifted and the section
was released to resume coal production.

9

SECTION 3

 Geology

The gas involved in the longwall inundations appears to be originating from the mine floor. It is
speculated that abutment stresses at the shields and pillars create floor heave and/or open floor
fractures which then act as pathways for the gas. How the gas pockets accumulate and what their
sources are is unknown, but information contained in Appendix 2.2-1 gives conditions which
might be factors.

A shale and sandstone stratigraphic sequence believed to be of marine origin exists under a small
leader seam below the Eagle seam. Marine shales can be a source for natural gas, but, in order
for the gas to be mobile, it must accumulate in open pore spaces or fractures in the rock. The
marine shale below the Eagle seam (the Betsie Shale) is capped by a sandstone unit that may act
as a reservoir for accumulating gas, and the lower shale units, or deeper shales, may be the source
beds for gas. The thickness of rock between the floor of the Little Eagle seam and the Eagle
seam itself is typically 10 to 15 feet. This interburden may act as an impermeable cap for gas, but
one that can be ruptured by abutment stresses during mining. Abutment stresses are related to
cover depth. The longwall inundations referenced above and the longwall face on April 5, 2010,
were in locations where the depth of cover exceeded 800 feet, the point at which abutment
stresses can challenge weakened rock.

The prior inundations have each occurred where the Little Eagle seam is comparatively thick.
Corehole data shows that the Little Eagle seam achieves its greatest thicknesses in localized
depositional channels which may be overlain with locally weak rock due to lithologic transitions
resulting in differential compaction slips and/or fractures that weaken rock and might provide
pathways for vertical migration of fluids and gases.

Structurally, the dominant extension fracture in the UBB mine roof trends approximately
N25°W. This is similar to the mine floor, where fracture orientations are more difficult to

10

SECTION 3

 ascertain, but which appear to be approximately N20°W. The dominant face cleat trends N80°E.
The coincident alignment of the roof and floor joint systems at UBB may facilitate the
occurrence of floor rupture and gas migration. This is not known for certain and requires further
study. Additional information pertaining to these conditions and factors can be found in
Appendix 2.2-1.

11

SECTION 3

 4

The Explosion

4.1

Operations in the 24 Hours Prior to the Explosion

4.2

The Explosion

4.3

Origin and Path of the Explosion

 4
4.1

The Explosion

Operations in the 24 Hours Prior to the Explosion

Pre-Shift Examinations for the Midnight Shift

On Sunday, April 4, 2010, UBB was idle in observance of Easter. UBB resumed operations late
Sunday night when midnight maintenance crews returned to work. Coal production at UBB,
however, did not restart until the dayshift on Monday, April 5, 2010.

Before midnight maintenance crews could enter UBB, the mine had to be pre-shifted.
Firebosses, who were to perform the pre-shift examinations of the mine prior to the Sunday
night midnight shift, began arriving at UBB shortly before 6 p.m. John Neely, John Skaggs and
John Bickford were to pre-shift the area of the mine inby Ellis Portal all the way to the longwall
and the two continuous miner sections known as #22 Tailgate and #22 Headgate. Bruce
Brackett and Terry Peterson pre-shifted the mine from the UBB Portal to the Ellis switch.

As part of that Sunday evening pre-shift examination, Neely, Skaggs and Bickford pre-shifted
the pumps at Ellis, the Glory Hole and North Mains. Their reports indicate no methane or
carbon monoxide were detected during their examinations of these areas, and that air movement
around the pumps was good and that they found no hazards.

1

SECTION 4

 Figure 1. Map showing location of portals and fans. See Explosion Region inset map, below.

Figure 2. Map of the Explosion Region.

2

SECTION 4

 Neely pre-shifted #22 Headgate; Skaggs pre-shifted #22 Tailgate and the longwall1; and
Bickford pre-shifted the Glory Hole and the Old North Mains areas of the mine. Neely‟s preshift examination report shows methane levels ranging from 0.1% to 0.25% in the face areas of
#22 Headgate with no carbon monoxide present and that air movement was adequate on #22
Headgate. The pre-shift examination report reflects that there were soft ribs in all three entries of
#22 Headgate and water was accumulating in the #3 face. The report also reflects that the haul
roads needed additional cleaning.

Skaggs conducted the pre-shift examination on #22 Tailgate. His report shows no methane or
carbon monoxide was detected in that area and air movement on the section was good. Skaggs‟
report indicated there was water up to his knees in the area in front of the section power center
over to the return.

Bickford pre-shifted the 7 North Construction Area of the mine. His report indicates that no
methane or carbon monoxide was found in this area, air movement was good and no hazards
were found.

Brackett pre-shifted the Barrier section and inby toward the Ellis Switch, while Peterson preshifted the Portal section area of the mine. Neither individual reported detecting any methane or
carbon monoxide in their pre-shift examination reports.

Since the midnight shift at UBB was a maintenance shift and did not produce coal, the belts did
not have to be pre-shifted, except in areas where workers were required to work on this shift.

The Midnight Shift

The midnight shift at UBB was a maintenance shift for all areas of the mine, which was the case
on April 5, 2010. The move crews on the midnight shift made splices on the belt at the #22
Headgate mother drive, moved belt on the #22 Headgate section, moved belt on the #22 Tailgate
1

During his interviews, John Skaggs acknowledged he did not fireboss the longwall face and reported that
fact to the longwall midnight foreman. See John Skaggs transcripts dated May 26, 2010 and March 1, 2011.

3

SECTION 4

 section and worked on the conveyor belt at the construction site just inby the Ellis Portal,
commonly referred to as Ellis Construction Site or LBB Construction Site.

The longwall maintenance crew changed the headgate cowl on the shearer, moved the monorail
down and worked on the longwall face conveyor chain replacing flights. In performing these
tasks, the crew cut bolts on the face chain and the cowl with a cutting torch, and tack welded the
nuts on the face chain and the nuts on the cowl. The longwall maintenance crew also changed
the oil and the bits on the shearer. The shearer was near the headgate while this work was
performed. The supply crew delivered the new cowl to the longwall and took the old cowl
outside.

The rock dust crew dusted the longwall belt and crosscuts from the mother drive toward the
longwall headgate. The rock dust crew utilized a pod duster that was pulled by a track mounted
locomotive. The farthest inby point of travel for the duster during that shift on the longwall track
was at approximately crosscut 17.

Pre-Shift Examinations for the Dayshift

Skaggs pre-shifted the LBB Construction Site for the dayshift. His report reflects that neither
methane nor carbon monoxide was found in this area of the mine, there were no hazards found
and air movement was good.

Neely pre-shifted the 7 North Construction Site for the dayshift. His report reflects that no
methane or carbon monoxide was found in that area, the air movement was good and no hazards
were found.

Bickford pre-shifted the pumps at Ellis, the Glory Hole and the North Mains for the dayshift.
His report reflects that he found no methane or carbon monoxide in these areas of the mine, air
movement was good and no hazards were found.

4

SECTION 4

 Skaggs pre-shifted the #4 Ellis, #5 Ellis and North #4 belts. No methane or carbon monoxide
was detected. Skaggs did report that the #4 Ellis and #5 Ellis belts needed rock dusted. He also
reported that the North #4 belt needed spot cleaned and rock dusted. Bickford pre-shifted the
North #5 and North #6 belts and reported that no methane or carbon monoxide was found. His
report also indicated that the North #5 belt needed spot cleaned from the #60 break to the flowthrough and that the North #6 belt needed spot cleaned and dusted.

Larry Brown pre-shifted the longwall for the dayshift. His report shows no methane or carbon
monoxide detected on the longwall and that there was adequate air movement with no hazards
observed. Brown also pre-shifted the longwall belt and found no methane or carbon monoxide,
and reported that the longwall belt needed spot cleaned.

Neely pre-shifted the North #7 belt, the Tailgate #1 belt and the Tailgate #2 belt and reported
that no methane or carbon monoxide was detected. Neely did report that the North #7 belt
needed rock dusted, the Tailgate #1 belt needed spot cleaned and spot dusted, and that the
Tailgate #2 belt had no hazards.

Neely conducted the pre-shift examination on #22 Headgate for the dayshift. He found 0.3%
methane in the #1 entry, and 0.1% methane in the #2 entry, the 2 right crosscut and the #3 entry.
No carbon monoxide was detected, and the report shows that air movement was adequate.
Neely also reported that the section needed cleaned and rock dusted.

Jason Thomas pre-shifted #22 Tailgate for the dayshift. He found no methane or carbon
monoxide and reported good air movement. Joe Coon pre-shifted the Barrier section for the
dayshift, found no methane or carbon monoxide and reported good air movement. He did,
however, report soft ribs on the section.

Brackett and Peterson conducted the pre-shift examinations on the belts on the south side of the
mine for the dayshift. They reported no methane or carbon monoxide. They did report areas on
the belts that needed cleaned and rock dusted.

5

SECTION 4

 The Dayshift

Coal production resumed at UBB during the dayshift on April 5, 2010. The first coal-producing
crews to enter the mine from the Ellis Portal were the #22 Headgate and the longwall crews,
who entered at approximately 6 a.m. At 6:38 a.m., the pump crew entered the mine from the
Ellis Portal. Two crews entered from the North Portal: The Barrier crew entered the mine at
approximately 6:10 a.m. followed by the #22 Tailgate crew at approximately 6:40 a.m.

The #22 Headgate, #22 Tailgate and Barrier section crews produced coal throughout the day.
The LLB Construction crew worked on cutting roof rock. Two men, Clifton Earls and James J.
Woods, were working on the south side at East Mains pulling rails at the time of the explosion;
however, they were in other areas of the mine earlier in the day performing other tasks.

Examinations of #21 Tailgate Conducted on April 5, 2010

During the dayshift on April 5, 2010, Jeremy Burghduff, foreman for the pump crew working in
the bleeder area behind the longwall, recorded 0.15% methane in the #4 entry at the 100 break
and 0.05% methane in the #3 entry at break 122.2 Burghduff reported water levels ranging from
9 to 36 inches in the area from break 85 to break 125. Burghduff also reported a small roof fall
at 52 break on Old 2 section on the track in the return and a damaged bolt at 83 ½ break.

Pre-Shift Examinations for the Evening Shift

Pat Hilbert received the pre-shift report for #22 Headgate from Dean Jones around 2:45 p.m.
Mr. Jones reported to Hilbert that he was at the feeder and found nothing unusual on #22
Headgate. Mr. Jones did report methane levels of 0.3% in the #1 entry, 0.2% in the #2 entry,
0.3% in the #3 entry and 0.3% methane in the #1 right scrap cut. Mr. Jones also reported that the
#1 entry needed rock dusted.
2

During the testing of all methane detectors, records indicate that the Solaris multi-gas detector assigned to
Jeremy Burghduff was not turned on during his entire shift on April 5, 2010. It was turned on during the
evening hours after the explosion.

6

SECTION 4

 Rick Lane called out the pre-shift report for the longwall evening shift to Kevin Medley at 2:40
p.m. Mr. Lane reported no methane, carbon monoxide or hazards, and that there was adequate
air flow on the longwall.

Michael Elswick pre-shifted the Glory Hole area and the belts on the North Glory Mains,
including the #6 and #7 belts. Mr. Elswick also pre-shifted the 7 North Construction Area for
the evening shift. He called this report outside at 2:30 p.m. His report reflects that he found no
methane or carbon monoxide in the areas, air movement was good and no hazards were found.

Mr. Elswick pre-shifted the pumps located in the North Mains and in the area of the Glory Hole;
Scott Halstead pre-shifted the pumps in the Ellis area of the mine. No methane or carbon
monoxide was detected at the pumps, air movement was good and no hazards were found.

Mr. Elswick pre-shifted the North #7 belt, the #22 Headgate belt, the Tailgate #1 and Tailgate #2
belts; no methane or carbon monoxide was detected. Mr. Elswick did report that the North #7
belt and the Tailgate #1 belt needed rock dusted, while the Tailgate #2 belt needed cleaned at the
tail. Mr. Elswick reported the #22 Headgate belt needed spot cleaned and rock dusted.

At 2:38 p.m., Steven Harrah called out the pre-shift examination on #22 Tailgate for the evening
shift. He reported no methane or carbon monoxide, good air movement and no hazards.
Bobby Baker pre-shifted the LBB Construction Site for the evening shift. His report reflects that
he found no methane or carbon monoxide in the work area, at the charger, at the pump or at the
overcast. He also reported no hazards.

Jim Bowyer conducted the pre-shift examination of the belts on the south side of the mine for
the evening shift. He reported no methane or carbon monoxide, but he did report that areas on
the belts needed cleaned and rock dusted.

7

SECTION 4

 State and Federal Inspections That Day

WVOMHS&T Inspector Wayne Wingrove arrived at the UBB Portal mine office at
approximately 7 a.m. to begin a quarterly inspection. After reviewing the mine map and the
examination books, Wingrove went to the Barrier section to begin his inspection. While on the
Barrier section that day he wrote three violations. One violation was for low air in the #2 entry.
The second violation was for an unsupported kettle bottom. The third violation was written for a
faulty manual disconnect on a scoop. Wingrove left the UBB mine at approximately 2 p.m. that
day.3

MSHA Inspector John Syner arrived at the UBB Portal mine office at 5:20 a.m. to begin an EO1
(quarterly) inspection. He held a pre-inspection meeting with Gary May and Rick Foster, and
gave a brief safety talk to the employees. He then went to the Barrier section to begin his
inspection. Mr. Syner found a cable splice that was not sealed to exclude moisture and the
section escapeway map was not up to date. A violation was issued for each. Syner found that
the rock dusting was good on the Barrier section. He left UBB between 1 p.m. and 2 p.m.4

The Longwall

The longwall crew entered the mine at approximately 6:04 a.m. When the crew arrived at the
longwall face the shearer was near the headgate. The first longwall call-out report occurred at
7:30 a.m. The 7:30 a.m. report indicates that they had made 0.2 passes, and the shearer was at
shield 70 headed towards the tail. The crew continued mining and arrived at the tail and
completed the shuffle between 8:30 a.m. and 9 a.m. At 10:30 a.m., the crew reported that the
shearer was at the head and that they were preparing to change bits. In the 11 a.m. call-out
report, the crew reported they replaced 25 bits and had lost the B-Lock on the tail ranging arm.
The B-Lock problem persisted until sometime between 1:30 p.m. and 2 p.m. In the 2 p.m. callout report, the crew reported that they had made an additional 0.2 passes and they were going to
re-check the B-Lock. At this time, the shearer was at shield 65 and headed toward the tail. In
3
4

Transcript of Wayne Wingrove dated May 10, 2010.
Testimony of John Syner dated May 14, 2010.

8

SECTION 4

 questioning why the shields had not been advanced and the panline not pushed on the headgate
side of the longwall face from around shield 60 to the headgate, production reports5 and
testimony 6 indicates that the crews had possibly been “swinging the face.” The 2:30 p.m. callout report states that the shearer was at shield 115 going toward the tail. At 2:40 p.m., Rick Lane
called out the pre-shift report for the evening shift to Kevin Medley. In that report, Lane stated
that he had detected no methane or carbon monoxide, no hazards and the air movement was
adequate. This is the last known communication received from the longwall.7

4.2

The Explosion

First Indications of the Explosion
At approximately 3:02.18 p.m. on April 5, 2010, the Pyott-Boone system computer, located on
the surface, recorded a communication loss from a CO monitor located at crosscut 99 on the 6
North conveyor belt (see Appendix 2.5-3 and 2.5-4). Due to the programming and data transfer
speed of the system, the event log entry could have been delayed by approximately 108
seconds.8

5

Production reports showing examples of mining extra footage from the tail or head the last few shifts on the
#21 Longwall face. See Appendix 4.1-1.
6
Kevin W. Medley testimony March 29, 2011
7
Seventeen (17) individuals invoked their 5th Amendment rights and were not interviewed as part of the
investigation process. These individuals may have knowledge of other communications from the longwall that
the State Investigation Team was unable to learn.
8
Pyott-Boone suggested the possible time delay, which would put the explosion earlier (3:00.30 pm)

9

SECTION 4

 All of the surface power for Ellis Portal was supplied via the Green High Voltage 12,470 volt
circuit, which originates at the main substation located at the UBB North Portal. A digital video
recorder (DVR), installed on the surface at the Ellis Portal, was powered by this circuit. As was
the mother drive (longwall belt drive), located at crosscut 102(see Figure 3). This circuit received
enough damage around crosscut 98 to de-energize the power. The internal clock on the DVR
stopped at 2:57.00 p.m. on April 5, 2010 due to power loss. After calculating time drift and
factoring in that the DVR records a time stamp in one minute intervals, the actual time is from
3:01.34 to 3:02.509 (see Figure 4).

The MSA Solaris Multi-Gas detector that was recovered from the longwall face at shield 82
recorded an over-range event for carbon monoxide and combustible gases sometime between
3:02.14 p.m. and 3:02.59 p.m. on April 5, 2010.10

The MSA Solaris Multi-Gas detector that was recovered from the #22 Tailgate mantrip recorded
an over-range event for carbon monoxide and combustible gases sometime between 3:01.34 p.m.
and 3:02.53 p.m. on April 5, 2010.11

The Joy Shearer has an internal data recorder referred to as a JAN 0 unit. A right handheld (tail
side) remote control caused the shearer to stop sometime between 2:59.32 p.m. and 2:59.38 p.m.,
after correction, on April 5, 2010.12

9

Times after time drift correction conducted at MSHA‟S Approval and Certification Center
Times after time drift correction conducted at MSHA‟S Approval and Certification Center
11
Times after time drift correction conducted at MSHA‟S Approval and Certification Center
12
Times after time drift correction conducted at MSHA‟S Approval and Certification Center
10

10

SECTION 4

 Figure 3. Location of events and damage summarized in the timeline below.

Figure 4. The timing of events around the time of the explosion (corrected for clock drift) expressed as a
range. NOTE: 3:02:18 p.m. is the first CO alarm, corrected for clock drift, and 3:00:30 includes the
possible delay in the Pyott-Boone system response. See Figure 3 for locations.

11

SECTION 4

 At approximately 2:40 p.m., the #22 Tailgate crew left their section and began traveling by a
battery powered, track-mounted mantrip to the surface. At approximately 2:58 p.m., James K.
Woods called the dispatcher for the road to Ellis switch and gave his location as 78 break. At 3
p.m., the longwall crew did not call outside. Shortly after 3 p.m., Greg Clay began to call the
longwall on the mine phone to get their 30-minute call-out report. He called two or three times
and received no response. He then heard a loud, strange noise coming from the fan.13 In his
interview, Clay stated “I guess about three minutes after 3:00, I just heard this bam (noise). I
thought the fan (north fan) had thrown a blade or something because it‟s making a real bad
noise.”14 He looked out the window and saw dust and debris coming out the North Portals. The
CO monitors were also alarming at this time.15 A short time later the fan returned to normal
operation. Brandon Davis stated in his interview that he was in the bathhouse at the North Portal
“and then I remember when I heard that fan (north fan) start making that noise. I looked and it
was two minutes after 3:00 on my watch.”16 Other eyewitnesses on the surface at the Ellis
Portal also reported strong wind, dust and debris coming out of the portal openings.

On the Ellis Portal side of the mountain, the evening shift longwall crew was at the mantrip a
few breaks underground at the time of the explosion. Crew members reported that at first they
felt a small breeze, but the air quickly picked up until it felt like what they described as hurricane
force winds carrying debris. Initially, crew members sought shelter in the crosscut and then
exited quickly out the Ellis Portal.

At the time of the explosion, Clifton Earls and James J. Woods were working on the south side
of the mine in the East Mains recovering rail when they heard something and stopped working.
Thinking they were just hearing things, they started back to work. At 4:10 p.m. they left their
work area to go outside. When they got outside they learned of the explosion. No one had
contacted them to inform them to evacuate.

13

Miners on the surface reported that the fan reversed; however, the State found no physical evidence that the
fan completely stalled or reversed.
14
Gregory Clay interview (June 23,2010)
15
Transcript of Adam Jenkins,(June 12, 2010)
16
Brandon Davis interview (September 1,2010)

12

SECTION 4

 Prior to the explosion, the dayshift LBB Construction crew had left their work area and were
traveling inby toward Ellis Switch on their way to the UBB North Portal when they experienced
intense pressure in their ears and then felt wind and debris hitting them. One member of the
LBB crew reported that the wind was so strong it pushed the mantrip outby toward the Ellis
Portal and that the foreman‟s gas detector began to alarm. During the explosion, something
caused the power on their mantrip to go off. While they sat on the track in the mantrip, the
foreman told his crew to get their self-contained self-rescue devices (SCSR) ready due to the
high level of CO he was picking up on his detector. The crew prepared their rescuers for
donning.17 The crew members deployed their SCSRs and some crew members donned them.
After the wind ceased blowing, the crew was able to energize the mantrip and immediately
began to tram toward the Ellis Portal where they made it safely outside.

Tim Blake worked on the #22 Tailgate crew. Mr. Blake reported that the #22 Tailgate crew quit
mining coal around 2:30 p.m. and began cleaning their section. At about 2:40 p.m. they left the
section on the mantrip. When they got to about 67 crosscut on the North Mains track, Mr. Blake
reported that all at once everything went black and it seemed like he was sitting in the middle of
a hurricane. Debris was flying through the air hitting the mantrip and the crew members. The
foreman‟s gas detector was alarming. Other members of the #22 Tailgate crew yelled for fellow
workers to don their rescuers. Mr. Blake held his breath while donning his rescuer; he reported
that the wind event lasted for approximately one to three minutes.

When things settled down, Mr. Blake got out of the mantrip and checked each crew member for
a pulse. He put SCSRs on each crew member except one whose rescuer he could not find.
When he finished working with each member of the crew, Mr. Blake looked at his watch; it was
3:57 p.m. he called on his radio for help, but received no response. His SCSR was running low
on air due to him attending to each individual miner on the mantrip. Mr. Blake stated that the
hardest thing he ever had to do was when at this time he left his crew to go for help. He walked
approximately 20 breaks and saw a mantrip approaching that had entered the mine from the Ellis
Portal.
17

Transcript of Bobby Baker, dated August 23, 2010.

13

SECTION 4

 Initial Actions after the Explosion

Workers had entered the mine preparing for the afternoon shift that started at 3 p.m. When the
forces from the explosion reached their locations near the drifts of the mine portals, it was
apparent that something huge had occurred inby. After the initial forces subsided, these workers
exited the mine.

Seconds after the explosion occurred, Dispatcher Adam Jenkins began calling the sections
underground to tell them to evacuate.18 About 15 minutes later, Jack Martin from the Barrier
section responded to the call.19 After receiving word to evacuate, Martin assembled all of his
men and began exiting the mine by mantrip through the UBB North Portal.20 Jenkins continued
to call for the working sections inby 78 break for at least another 30 minutes after the explosion;
however, there was no response.21

From the Ellis Portal

About 20 to 25 minutes after the effects of the explosion stopped coming out of the Ellis Portal,
Chris Blanchard, Jason Whitehead, Jack Roles, Everett Hager, Wayne Persinger and Pat Hilbert
entered the mine on a mantrip through the Ellis Portal. Hilbert was driving the mantrip, and
when they got to the first overcast they stopped to clear cinder blocks from the track. When they
arrived at the Ellis Switch, they called outside to report their position and proceeded inby. At
approximately 47 break, they saw a single light walking toward them that was Mr. Blake. He
told them his crew was about 20 breaks inby. Blanchard, Whitehead, Roles, Persinger and
Hager left Hilbert and Mr. Blake with the mantrip and proceeded on foot inby. About ten
minutes later, Roles ran back and said they needed the mantrip. Roles stayed with Mr. Blake
while Hilbert took the mantrip inby.
From the UBB North Portal

18
19
20
21

Testimony of Adam Jenkins dated June 12, 2010.
Testimony of Adam Jenkins dated June 12, 2010.
Testimony of Jack Martin dated July 20, 2010.
Testimony of Adam Jenkins dated June 12, 2010.

14

SECTION 4

 Shortly after the explosion, Gary May entered the mine on foot from the UBB North Portal.
Rick Foster then entered the mine through the UBB North Portal on a mantrip. Berman Cornett
and James Walker were the next people to enter the mine through the UBB North Portal, doing
so on foot. They met Foster who was removing pins from the double airlock doors just inby the
portal. The three men together continued inby on the mantrip. They switched out at the Plumley
Switch to allow the Barrier section crew to exit the mine. After the Barrier section crew passed,
Foster‟s mantrip proceeded inby to the Ellis Switch. It was at this location where they met Gary
May who had traveled to this point on foot. Together they proceeded inby on North Mains
towards 78 break. At approximately 47 break they saw lights coming down the track. The
lights belonged to Jack Roles and Mr. Blake who were walking outby.22
Performance Coal Company’s Initial Rescue and Recovery Efforts
When Hilbert arrived at 66 break he saw the mantrip with the victims.23 They loaded Steve
Harrah, James Woods, Bill Lynch and Carl Acord on Hilbert‟s mantrip. Wayne Persinger rode
out with Hilbert giving CPR to Mr. Woods. About two or three breaks outby, Hilbert met Gary
May, James Walker and Berman Cornett. These three boarded Hilbert‟s mantrip to perform
CPR on two of the other victims as they proceeded outside along with Rick Foster, who was
driving the mantrip brought in from the North Portal. Blanchard, Whitehead and Hager loaded
the other victims on the second mantrip and Hager brought them outside. After all the #22
Tailgate crew members had been loaded on the two mantrips and started towards the surface,
Blanchard and Whitehead continued inby.24
Massey‟s Southern West Virginia Mine Rescue team is located near the main guard shack at
Performance Coal Company. Rob Asbury, Mine Rescue Coordinator for Massey‟s southern
team, received the emergency call shortly after the explosion. Asbury was informed to take the

22
23
24

Transcript of James Walker dated August 17, 2010.
Testimony of Pat Hilbert dated July 10, 2010.
Testimony of Pat Hilbert dated July 10, 2010.

15

SECTION 4

 mine rescue truck and rescue team members to the Ellis Portal. After arriving at the Ellis Portal
and benching their apparatus, the team began to enter the mine.

4.3

Origin and Path of the Explosion

Mine explosions have an ignition point and an explosion origin. While the location for both can
be the same, much depends on the distribution of combustible fuel at the ignition point.
Ignitions of fuel-rich gas bodies may not explode upon ignition but burn around their perimeter
until the right mixture is found and then explode. Because coal dust must be in suspension for a
dust explosion to begin, methane was the fuel for the ignition. Depending on the fuel
distribution, the mine geometry and other factors, the explosion can take one path or multiple
paths. Part of the job of investigators is to discover the evidence to make these interpretations
about the origin and path of the explosion.
The origin of the explosion was determined initially by a process of elimination. Locations
where explosion forces first propagated into, then propagated out of, cannot be where the
explosion originated. Using the various tools and methods described in the report appendices,
the sequence and direction of explosion wind forces can be established. Some of the essential
information is summarized in Figure 5. Based on this information, candidate locations for the
explosion origin were reduced to just a few possibilities. It was determined that the Tailgate
region of the longwall was where the explosion began. Subsequent work narrowed the location
further.

16

SECTION 4

 Figure 5. There are multiple directions of wind forces during a mine explosion. Using indicators of initial
wind forces and final wind forces the explosion candidate locations for origin can be inferred. See
Section 7.9, Map 15 for larger map.

Investigators concluded that the explosion likely occurred in or on the rubble (gob) behind the
shields at a location near the Tailgate (see Figure 6). The fuel for the ignition was methane that
was mostly liberated behind the shields, either from floor cracks and/or by expansion of methane
from the gob due to falling barometer and the ventilation gradient. It is believed the ventilating
pressure gradient25 and flow direction moved the methane along the gob until it reached the
intersection of #7 entry at shield 176. Behind this shield is the corner of the gob; the gob
extended across #7 entry creating blockage in that entry, restricting ventilation. Upon ignition,
the gas could have burned as a wick into the gob before explosion generating forces occurred
nearby. After burning for a time, a methane explosion occurred in the gob near the tailgate,
which then propagated along multiple initial paths. The explosion dynamics due to gas
explosion in and on the rubble would be highly turbulent and complex. A methane flame will
25

See Section 7.3

17

SECTION 4

 propagate through cracks that are at least 0.1 inch wide but the flame and gases cannot travel and
expand freely, so in smaller openings flame propagation will be slower than in open air. The
explosion front may be compressed due to multiple reflections on rock surfaces especially if the
gas body and open spaces between the rubble are large, which would be the case near the edges
of the rubble. Explosion of pre-compressed gases, called pressure piling, could create higher
explosion pressures, and the explosion could have exited the gob with some force.
The explosion event gives evidence that it took multiple paths. One explosion path was along
the gob fringe behind the shields (A), a second path was into the T-split area of the longwall (B),
and a third split exited the gob approximately 300 to 400 feet west of the longwall shearer (C),
(see Figure 6, Appendix 7.9-5).
Once in the #21 Tailgate, the methane explosion transitioned to a coal dust explosion
propagating eastward in the #7 entry away from the “T-split”, also southward, through the first
crosscut outby the shearer and also westward from where explosion forces exited the gob a few
crosscuts west of the longwall. The early stages of the dust explosion may have involved a
hybrid fuel of methane and dust. As the explosion continued to propagate, the composition of
fuel is believed to be coal dust generated from the natural rib spalling from principally the Top
Coal bench and also float dust from mining operations (see Appendix 8.2-1).

Figure 6: Evidence suggests the explosion began in the gob behind the longwall shields. RED arrows
indicate the inferred path of the initial explosion. BLACK arrows show the initial paths of propagation
once these forces found their way into #21 Tailgate. The actual ignition may or may not have occurred
at the explosion origin.

18

SECTION 4

 Information contained in the West Virginia Flames and Forces Map, various compilations of
summary maps, and other information were used to determine the likely initial path of the
explosion through the mine. This appears in Figure 7.
Initial Path of the Explosion
The direction and sequence of explosion propagation was determined by mapping and analyzing
bent roof pans (see Appendix 7.9-2 ), and determining the breach directions of ventilation
stoppings, conveyor belt structures, waterlines, airlock doors, etc. (see Appendix 7.9-4 ), and
also by compiling and analyzing other data such as the geometry of impacted dust on roof bolt
heads (see Appendix 7.9-1 ). Final forces were determined by documenting the resting
positions of lightweight paper, plastic, cloth, etc., and supplemented by “impacted „V‟ dust
deposits” on standing cylindrical structures (see Appendix 7.9-3). Explosion forces in the gob
exited the tailgate entries, and the general direction of the explosion could be traced as it moved
out of the tailgate entries.

Initial explosion forces in #21 Tailgate propagated westward toward the Bandytown fan and
also eastward toward the #21 Cross-over. Eastward, the leading edge occupied the #7 entry
until the #21 Cross-over was reached, and from there the flame speed in the southern entries
increased relative to the northern entries, producing high pressure damage to roof pans and
resulting in the #1 entry assuming the leading edge. East of the #21 Cross-over the leading edge
remained in the #1 entry of #21 Tailgate for the remainder of its initial propagation to
approximately the switch near break 78 , where propagation subsided to extinction.
A branch into the #21 Cross-over propagated north and then branched eastward into the
southern entries of the Cut-across, continuing eastward. These wind forces necked down to just
three entries when they merged with the North Glory Mains, resulting in increased pressures and
severe damage to roof pans and belt structures in the crosscuts of North Glory Mains at this
confluence. The explosion quickly subsided to extinction to the northwest of the confluence.
Southeast of the confluence the explosion leading edge was in the south entries, and propagation
continued until reaching the 1 North Mains and Parallel Mains where the explosion quickly
subsided to extinction.

19

SECTION 4

 Figure 7: After the explosion in the longwall gob exited into the #21 Tailgate, its initial path continued to
find fuel and propagated along a path indicated with RED ARROWS. Indicated with GRAY ARROWS
are locations where the initial explosion forces subside, due to lack of fuel or other factors.

The branch of the explosion that continued north in the #21 Cross-over branched three ways at
#21 Headgate. The east branch propagated quickly to extinction while the north branch crossed
the #21 Headgate and entered North Glory Mains, where it propagated approximately 1,000 feet
northeast then subsided. The west branch in #21 Headgate entered the belt entry26 first but
within a few hundred feet the opposite side (#4 and #5 entries) assumed the leading edge of the
explosion, when flame velocity increased in the last 500 feet before the confluence with the #22
Cross-over. Pressures were high in this confluence, as evidenced by severe roof pan damage.
One branch of the explosion then proceeded north into the #22 Cross-over and the remainder
continued west into #22 Tailgate and the three remaining entries of the #21 Headgate. Forces
also entered the longwall and subsided to extinction.
West of the #22 Cross-over, the forces that arrived at the three-entry #21 Headgate entered first
at a crosscut near the mouth of #22 Tailgate and entered #21 Headgate through a water-filled
26

The southern entry (or #1 entry)

20

SECTION 4

 dip. The explosion was quenched by this water barrier, quickly subsided to extinction and did
not propagate further west in #21 Headgate (see Appendix 4.3-1).

The branch of the explosion that continued north in the #22 Cross-over appears to have lost
some momentum as it crossed the mouth of #22 Headgate and branched east and west, and
continued north. It subsided to extinction in the last few hundred feet of the #22 Cross-over after
crossing the #22 Headgate. Upon making the turn west in the #22 Headgate (a three entry
heading) the leading edge of the explosion entered the belt entry (the #1 entry), but within 1,000
feet the #3 entry had accelerated and caught up with #1 entry resulting in severe pan bending in
the #1 and #2 entries. At the dead end of #22 Headgate high reflected pressures developed in the
#2 and #3 entries.

The explosion forces from #22 Cross-over also turned southeast and into the North Glory Mains,
while another branch propagated north into #8 North and entered the Glory Hole area from a
second location. Along the southeast solid rib of the Glory Hole area is evidence of reflected
pressure and severe pan bending from these forces propagating through the Glory Hole area.
The propagation of forces into #8 North (North Jarrell‟s Mains) gathered momentum as the
explosion traveled north, producing severe damage to roof pans approximately 600 feet south of
the #9 North (West Jarrell‟s Mains) headings. This damage was primarily along the outer
entries. After making the turn west into #9 North, the leading edge of the westward propagation
started in the southern entries; however the northern entries gained speed and developed into the
leading edge of the explosion during the last 1,000 feet of propagation to the #9 North western
dead end. Part of the dead end was flooded, so some information is unavailable, but forces at the
dead end were determined to be north-to-south and they produced severe pan damage from
reflected pressures in the southwest corner of #9 North.

21

SECTION 4

 Return Forces Sufficient to Bend Roof Pans

After the passage of the initial pressure, subsequent wind forces are generated by a variety of
factors involving reflected pressures, cooling of combustion gases in the seconds after the
explosion and, we believe in some cases, additional fuel consumption and propagation. The
source of oxygen in this explosion was unusual in that the airflow to Bandytown fan was a
straight run through in the #21 Tailgate and #21 Headgate entries, independent of ventilation
controls. The UBB ventilation fans remained operational after the explosion, so that even
though stoppings were destroyed, partial27 ventilation resumed immediately after the explosion
reverberations stopped.

Figure 8: Return explosion forces of sufficient strength to bend roof pans are indicated with BLUE
ARROWS. Indicated with GRAY ARROWS are locations where the initial explosion forces subside to
extinction.

A sense of the relative strength of these secondary forces comes from the bending of roof pans.
A summary of the subsequent bending of roof pans appears in Figure 8. The blue arrows show
where explosion return forces were strong enough to bend roof pans. The gray arrows indicate
27

Internal ventilation controls were destroyed

22

SECTION 4

 areas where return forces are too weak to bend roof pans, but their presence is known due to the
Final Forces indicators observed and documented (see Appendix 7.9-3).

23

SECTION 4

 5

The Rescue and Recovery

5.1

Chronology of Events

5.2

List of Victims

5.3

Monitoring Conditions in the Mine

 5

The Rescue and Recovery

5.1

Chronology of Events

April 5, 2010
At approximately 3:02 p.m., an explosion occurred at the UBB mine located at Montcoal
(Raleigh County), West Virginia. Within minutes members of the Massey Energy mine rescue
teams were notified.

Shortly after the explosion, 10 employees from UBB entered the mine. One group of six men,
Patrick Hilbert, Chris Blanchard, Jason Whitehead, Jack Roles, Everett Hager and Wayne
Persinger, entered from the Ellis Portal on a battery driven mantrip on rail.1 The other group
consisting of four men, Rick Foster, Gary May, James Walker and Berman Cornett, entered
the mine from the North Portal.2

Timothy Blake, one of the two survivors riding on the #22 Tailgate section mantrip that was in
the explosion area, began exiting the mine on foot.3 4 He traveled approximately 20 crosscuts
outby the stopping point of the mantrip when he saw lights of a mantrip coming in the mine.
This was the mantrip that had entered from the Ellis Portal. Upon reaching Mr. Blake, the
1

Patrick Hilbert interview July 10, 2010, page 57
James Walker interview August 17, 2010, page 60
3
Timothy Blake interview September 30, 2010, page 44
4
The distance from crosscut 66 to Ellis Portal is 11,600 feet
2

1

SECTION 5

 mantrip stopped. One employee, Patrick Hilbert, remained with the survivor and the mantrip.
The others proceeded on foot inby near crosscut 67 where they located the other eight miners.
After finding the eight miners, Mr. Roles ran back to the mantrip where Mr. Blake and Mr.
Hilbert were located and stated that the mantrip at their location was needed inby. Mr. Hilbert
then preceded inby with the mantrip and Mr. Roles remained with Mr. Blake. When Mr.
Hilbert reached the area where the original mantrip was located, the eight miners were then
loaded on the two mantrips and driven to the surface. Chris Blanchard and Jason Whitehead
continued to advance into the mine. After traveling toward the surface a short distance, those
exiting the mine were intercepted by men who had entered from the North Portal. They
boarded the mantrips with the victims and all three mantrips traveled to the Ellis Portal. These
mantrips arrived at the Ellis Portal at approximately 4:45 p.m.5

A short time after the nine miners who were on the mantrip near crosscut 67 were brought to
the surface, a three-man mine rescue team from the Massey Southern West Virginia team
entered the mine from the Ellis Portal. Two of the team members were on foot traveling in
front of the battery powered mantrip checking the atmosphere with a handheld detector and
repairing the telephone line in order to have communications to the command center. The
three team members traveled to North Mains, crosscut 78, where rail travel was no longer
possible due to destruction and obstacles on the track. All three then started exploration on
foot. While traveling alongside of the conveyor belt, the team encountered a break in the fresh
water supply line. They retreated back to crosscut 42 to turn the water off. A second threeman mine rescue team from the Massey Southern West Virginia team entered the mine at 6
p.m. from the Ellis Portal and traveled by a battery powered rail mantrip to crosscut 78, where
they met with the first three-man team.6

Both Massey Southern West Virginia teams left crosscut 78 and traveled a short distance in
the North Glory Mains toward the longwall mother drive. They saw Mr. Whitehead and Mr.
Blanchard walking toward them in the track entry of the North Glory Mains.7
5

Whitesville Vol. Fire Department timeline
Massey management command center notes, page 9
7
Shane McPherson interview March 9, 2011, page 29
6

2

SECTION 5

 The WVOMHS&T Mine Rescue Teams were notified of the event after completing an
afternoon of training at the Southern West Virginia Community and Technical College at
Logan, West Virginia.

The teams were deployed and arrived at the Ellis Portal at

approximately 6:05 p.m. After a brief session with mine management, the teams were sent to
the Upper Big Branch North Portal.

At approximately 7:42 p.m., two Massey Energy teams, Knox Creek and East Kentucky,
entered the mine at the North Portal. Two WVOMHS&T State Inspectors, Danny Cook and
Greg Norman, members of the WVOMHS&T Mine Emergency Team, accompanied by Scott
Kingery and Jason Campbell, members of the Massey Southern West Virginia mine rescue
team, also traveled with this group. Three MSHA Inspectors, Gerald Cook and Mike Hicks,
who are members of the MSHA Mine Emergency Unit, and Fred Wills, who was not an
apparatus wearer at the time, also traveled with this group by rail to crosscut 78. From that
point they traveled on foot to the next fresh air base, which was established at crosscut 84.

Members of the Massey Southern West Virginia and East Kentucky teams began exploring
the track and conveyor belt entries inby towards the #21 Longwall section. Six miners were
located during these initial explorations. Victim #18 was located between the track and belt
entry at the mother drive starter box.9 Victims #2,10 #3,11 #4,12 and #513 were in the track
entry and Victim #614 was located at the longwall stage loader.15

Two members of the Massey Southern West Virginia team and Gerald Cook arrived at the
longwall stage loader. The two Massey team members began to explore across the longwall

8

Later positive identified as Michael Elswick
Shane McPherson interview March 9, 2011, page 31
10
Later positive identified as Adam Morgan
11
Later positive identified as Cory Davis
12
Later positive identified as Joshua Napper
13
Later positive identified as Charles “Timmy” Davis
14
Later positive identified as Rex Mullins
15
Shane McPherson interview March 9, 2011, page 33
9

3

SECTION 5

 face. Mr. Cook remained near the headgate of the longwall to relay communications from the
two Massey team members to the fresh air base.

As they started across the face, their detector was reading 20.8% oxygen, zero methane and 20
parts per million of carbon monoxide. Approximately halfway across the face, the carbon
monoxide went from 20 to 40 parts per million, and somewhere between shield 120 and 125
the detector showed 60 parts per million of carbon monoxide and 2.0% methane.16 Further
travel down the longwall face was prohibited due to lack of communications. During the
exploration, six deceased miners (Victims #7,17 #8,18 #9,19 #10,20 #1121 and #1222) were
located on the longwall face.23

At approximately 8:36 p.m., the WVOMHS&T South Team, with Eugene White as captain,
was instructed to enter and explore the area known as the cross-over or longwall pull-off area.
The WVOMHS&T North Team, with Barry Fletcher as captain, remained on standby at the
fresh air base for the West Virginia South Team. Since the West Virginia South Team had
never been in the area prior to the explosion, Chris Blanchard, a member of the UBB
management team who had been located at the crosscut 78 fresh air base, accompanied them.
There were foot prints in this area that were later to be determined as the footprints of Mr.
Blanchard and Mr. Whitehead, who had been in the area after the explosion. The teams
encountered damage to the ventilation controls and heavy soot in this area. The team
members had apparatuses with them but, due to the air quality, had not gone under oxygen.
The temperature of the atmosphere was hot. Team members exploring in this area of the mine
noted that their eyes were burning.24 The team returned to the fresh air base at 11:42 p.m.

16

Shane McPherson interview March 9, 2011, page 34
Later positive identified as Richard Lane
18
Later positive identified as Grover Skeens
19
Later positive identified as Joel Price
20
Later positive identified as Gary Quarles
21
Later positive identified as Christopher Bell Sr.
22
Later positive identified as Dillard Persinger
23
Shane McPherson interview March 9, 2011, page 34
24
Eugene White interview May 20,2010, page 28
17

4

SECTION 5

 At 9:45 p.m., two additional mine rescue teams were deployed from the staging area. The
teams were the WVOMHS&T Welch Team, with Bill Tucker as captain, and the Beckley ICG
Team. The teams traveled to crosscut 78 then proceeded on foot to the next fresh air base at
11:42 p.m. At about crosscut 86 on 6 North Belt they discovered a small hot spot that was
quickly extinguished.

The Knox Creek team was instructed to explore up to the Glory Hole. They made it to the
Glory Hole belt head area, which was near crosscut 127-128. They were under oxygen. Their
detectors indicated that they had from 80 to 100 parts per million of carbon monoxide; at one
time, 1,500 to 1,600 parts per million of carbon monoxide was detected around the Glory Hole
belt head. Their apparatuses began to get low on oxygen. They were relieved out by East
Kentucky #2 team and a couple of team members from the Massey Energy Southern West
Virginia team along with State mine rescue team members and MSHA Mine Emergency Unit
team members. The Knox Creek team was then instructed to explore back to crosscut 78 by
way of the # 5 entry in search of anything or anyone.25

At 10:55 p.m., the Massey Energy East Kentucky #2 team and the WVOMHS&T North team
began advancing from the fresh air base towards the #22 Headgate continuous miner section.
At 11:11 p.m., the Massey Energy Southern West Virginia team had located the emergency
shelter on the longwall. The shelter had not been deployed. At 11:20 p.m., the East Kentucky
#2 team went under oxygen. At 11:27 p.m., the team called its location as being three breaks
inby the continuous miner panel of #22 Headgate. They had 14.7% oxygen, 8,676 parts per
million of carbon monoxide and 3.3% methane. At 11:37 p.m., the team had advanced one
crosscut and called out heavy smoke, 7.4% oxygen, 5,388 parts per million of carbon
monoxide, 0% methane, and the temperature of 80 degrees F. At 11:55 p.m., the East
Kentucky #2 team found a mantrip on the track near crosscut 19 on the #22 Headgate section
between survey station 24394 and 24401. Six victims (#14,26 #15,27 #16,28 #17,29 #1830 and
#1931) were on the mantrip. All were deceased. No SCSRs had been deployed.
25

Scott Kingery interview 1/5/2011, pages 22 & 24
Later positive identified as Ricky Workman
27
Later positive identified as Howard “Boone” Payne Jr.
26

5

SECTION 5

 April 6, 2010
At 12:16 a.m., the East Kentucky #2 team reported from just inby the mantrip that they had
located on the track an air quality reading of 3.2% oxygen, 9,999 parts per million of carbon
monoxide and over-range on methane. At 12:37 a.m., the East Kentucky team reported heavy
smoke, that they could not determine air flow direction and that all detectors showed overrange on methane and carbon monoxide, with 3.2% oxygen. At 12:45 a.m., all rescue teams
were ordered by the command center to retreat out of the mine due to the explosive mixture of
gas and evidence of a fire.32

April 7, 2010
There were no rescue or recovery activities this day due to the conditions found on April 6,
2010, when the teams were exploring. Boreholes were being drilled from the surface into the
mine. The boreholes were to be used to collect atmospheric samples, and at a later date, to
inject nitrogen into the mine.

April 8, 2010
At 1:45 a.m., samples collected from the boreholes permitted reentry into the mine for rescue
operations.

At 4:55 a.m. four mine rescue teams entered the mine. A member of the WVOMHS&T Mine
Emergency Team and a member of the MSHA Mine Emergency Unit traveled with each team
from this point forward.

All teams going underground continued to take supplies such as first-aid supplies, EMT kits,
AEDs, CARE-vents, basket stretchers, ventilation curtain, pogo-sticks, telephones, telephone
line and victim recovery apparatus that would be needed in the rescue and recovery.
28

Later positive identified as Ronald Maynor
Later positive identified as James “Eddie” Mooney
30
Later positive identified as Kenneth Chapman
31
Later positive identified as William Griffith
32
State command center notes 4/6/2010, pages 13 & 14
29

6

SECTION 5

 As the teams traveled in the mine from the North Portal, one team member traveled on foot
ahead of the mantrips to check the air quality and look for hazards. The teams stopped at
crosscut 42 of the Old North Mains to check the seal (set 14). The seal was intact, water in the
trap was empty and air was out gassing. Air quality was 16.6% oxygen, 0.25% methane and
no carbon monoxide. The team continued to advance toward the longwall mule train with the
mission of establishing a fresh air base.33

At 9:29 a.m., the command center received analysis of air samples taken from a borehole
on the surface indicating that there was an explosive mixture of gas in the mine. The
mine rescue teams were ordered to evacuate to the surface immediately.34 At 10:55 a.m.,
all rescue teams had returned to the surface.

April 9, 2010
At 12:42 a.m., two mine rescue teams reentered the mine traveling by rail on a battery
powered mantrip to crosscut 78. After reaching crosscut 78 they traveled on foot to the fresh
air base that had been established near the longwall mule train. The teams advanced inby and
established a fresh air base at Headgate 1 North (a.k.a. #21 Headgate), #2 entry between
crosscuts 26 and 27. One team remained at this location as a back-up team, while the other
team explored under oxygen. The team explored in the area at the mouth of #22 Tailgate
where they encountered light smoke. When the emergency shelter was located it was
observed that the shelter had not been deployed or damaged. The team retreated back to the
fresh air base.

At 4:44 a.m., the command center received analysis of air samples taken from a borehole on
the surface indicating that there was evidence of a fire and that a methane air mixture was
close to explosive range. Teams were ordered to retreat to the surface immediately.35 At 6:11
a.m., all rescue teams had returned to the surface.
33

Massey management command center notes 4/8/2010, page 5
State command center notes 4/8/2010, page 3
35
MSHA command center notes 4/9/2010, page 3
34

7

SECTION 5

 At 4:15 p.m., two teams reentered the mine traveling by rail on a battery powered mantrip to
crosscut 78. At this point, they traveled on foot to the fresh air base that had been established
in the Headgate 1 North #2 entry between crosscut 26 and 27. From there they traveled the
#22 cross-over entries toward the #22 Headgate section. One team remained near the mouth
of #22 Headgate section while the other team explored inby towards the section. Team
members were placed along the travel route to maintain radio communications back to the
fresh air base.

At 7:28 p.m., two teams, along with members of Task Force 1, reentered the mine on a battery
powered rail mantrip traveling to crosscut 78. At this point, they traveled on foot to the fresh
air base that had been established in the Headgate 1 North #2 entry between crosscut 26 and
27. One team remained at the fresh air base while the other team began to advance across the
longwall face. As the team advanced across the longwall face, team members were placed at
locations to permit radio communications back to the fresh air base.

At 8:10 p.m., while exploring the #22 Headgate section, a rescue team under apparatus
discovered Victim #2036 near survey station 24446. At 8:16 p.m., near survey station 24447,
Victim #2137 was found. At 8:20 p.m. Victim #2238 was found near survey station 24458. All
were located in the track entry.39

At 11:20 p.m., while exploring across the longwall face, a rescue team member discovered
Victim #1340 in the #3 gate shield.
At this time, all missing miners had been found and preparation was made for recovery.41

36

Later positive identified as Joe Marcum
Later positive identified as Gregory Brock
38
Later positive identified as Edward “Dean” Jones
39
State command center notes 4/9/2010, page 5
40
Later positive identified as Nicolas McCroskey
41
See Appendix 5.1-1, list of victims (Victim Informational DataSheet).
37

8

SECTION 5

 April 10, 2010
At 12:15 p.m., there were four rescue teams along with Task Force 1 underground. Three
teams proceeded to the surface while one team and Task Force 1 began preparing the victims
along the longwall track and belt for transportation. After completing this task, they proceeded
out of the mine, arriving on the surface at 2:37 a.m.

At 3:40 a.m., a battery powered rail locomotive with a rubber tired four wheeler secured on a
lowboy entered the mine traveling to crosscut 78, where it was unloaded. The locomotive and
lowboy then returned to the surface.

At 3:55 a.m., three mine rescue teams entered the mine and advanced to crosscut 78. The
teams then traveled to the area of the longwall track that had been designated as a staging area,
and continued to prepare the victims for transport.

At 6:46 a.m., another rubber tired four wheeled battery vehicle was hauled underground and
unloaded at crosscut 78. Six of the deceased miners that had been brought to the staging area
were then transported to crosscut 78 by a rubber tired trailer pulled by the battery powered four
wheeler. These victims were then transported to the surface.

At 9:45 a.m., the mantrips with Victims #1-6 left crosscut 78 arriving on the surface at 10:15
a.m.42

At 1:59 a.m., a mantrip with Victims #8, #11 and #13 arrived on the surface, along with the
recovery teams.43

At 3:58 p.m., mine rescue teams entered the mine and traveled to the longwall where they
started bringing victims off the longwall face to the staging area near crosscut 78.
42
43

MSHA command center notes 4/10/2010, page 9
MSHA command center notes 4/10/2010, page 10

9

SECTION 5

 At 9:15 p.m., a mine rescue team left crosscut 78 for the surface. Two deceased miners were
being transported at this time.

At 10:11 p.m., Victims #10 and #13 reached the surface.

At 10:30 p.m., two mine rescue teams left the surface traveling to crosscut 78. From crosscut
78 they traveled on foot to the 7 North Area.

At 11:30 p.m., two mine rescue teams left the surface traveling to crosscut 78. From crosscut
78 they traveled on foot to the staging area on the longwall track.

April 11, 2010
At 12:12 a.m., two rubber tired four wheeled battery vehicles with trailers on lowboys pulled
by three locomotives entered the mine and traveled to crosscut 78.

At 12:35 a.m., two mine rescue teams proceeded to the 8 North/9 North Area.

At 1:30 a.m., the team exploring the 8 North Area went under oxygen. They called out
atmosphere qualities of 20.8% oxygen, 36 parts per million of carbon monoxide and 0%
methane. The team then called out atmosphere qualities at crosscut 149 with 19.8% oxygen,
2,151 parts per million of carbon monoxide and 1.3% methane. They continued to explore
into the 9 North Area.

At 2:10 a.m., a rescue team at crosscut 78 retrieved the four-wheelers and trailers and traveled
back to the fresh air base along the longwall track in preparation to start recovering the victims
from this area.

At 2:24 a.m., the rescue team exploring the 9 North panel informs the command center that all
ventilation controls are missing.

10

SECTION 5

 At 3:10 a.m., the team exploring the 9 North Area retreated back to the fresh air base.
At 3:30 a.m., a “mapping crew” consisting of State, MSHA and Massey personnel entered the
mine. The mapping crew made a map of the areas where temporary ventilation controls were
going to be installed to ventilate the mine to recover the victims in the #22 Headgate Area.

At 4 a.m., a mantrip with Victims #7 and #9 left crosscut 78 traveling to the surface.

At 4:58 a.m., a rescue team explored the #22 Tailgate panel while a back-up team remained at
the longwall fresh air base. Two crosscuts into the panel, the team had atmosphere readings of
20.8% oxygen, 120 parts per million of carbon monoxide and 0.02% methane. Four crosscuts
into the panel, the team had atmosphere readings of 20.4% oxygen, 225 parts per million of
carbon monoxide and 0.04% methane.

At 5 a.m., a mantrip with Victims #7 and #9 arrived on the surface.

At 8:41 a.m., two mine rescue teams entered the mine. When they arrived at the fresh air base
at crosscut 78 they began to carry supplies further into the mine.

At 11:20 a.m., the mapping crew returned to the surface.

At 2:24 p.m., all ventilation work at the mouth of Tailgate 1 North, the pull-off area and
Headgate 1 North had been completed with curtains installed, but rolled up. This area was
noted on a modification plan/map as areas A, B, C and D.

At 2:35 p.m., two mine rescue teams entered the mine.

At 3:07 p.m., nitrogen was being injected from the surface into #22 Headgate section.

11

SECTION 5

 At 5:01 p.m., motor crews were on the surface. A total of 35 mine rescue team members were
underground.44

At 5:16 p.m., a team entered the mouth of #22 Headgate, while under oxygen, with a back-up
team at the fresh air base at crosscut 129.

At 6:07 p.m., instructions were given by the command center to move the fresh air base from
crosscut 78 to crosscut 129. It was determined that the walking travel time between the two
fresh air bases was 45 minutes.45

At 10:03 p.m., all ventilation curtains that had been installed at the pull-off area and the
Headgate 1 North Area (this area is noted as B – C – D on a modification plan/map) were
dropped and secured.

At 10:30 p.m., ventilation curtains at the mouth of Tailgate 1 North and entries 3 through 7
(this area is noted as A on a modification plan/map) were dropped and secured. Teams in the
area were instructed to check #1 and #2 entries to make sure that the curtains had been
installed across the entries. At 11:38 p.m., all mine rescue team members were on the surface.
No one remained underground.46

April 12, 2010
At 5:02 a.m., the command center made the decision to send in four mine rescue teams.

At 6:50 a.m., four mine rescue teams were sent underground.

At 9:15 a.m., two mine rescue teams were sent underground.

44

State command center notes 4/11/2010, page 32
State command center notes 4/11/2010, page 33
46
State command center notes 4/11/2010, page 35
45

12

SECTION 5

 At 10:12 a.m., a rescue team was instructed to install checks across #4, #5, #6 entries at
crosscut 129, directing the air up to the Glory Hole and then across the temporary seals at
crosscut 140 in the 8 North Area.47

At 11:14 a.m., the rescue teams were instructed to start the re-ventilation process of #22
Headgate by installing curtains across the cut-through panel between #22 Headgate and #22
Tailgate.

At 11:19 a.m., three members of the WVOMHS&T team entered the mine accompanied by
MSHA and Task Force 1. This group traveled by mantrip to crosscut 78 and then on foot to
the fresh air base at crosscut 129 to remove the victims from the # 22 Headgate mantrip and
prepare them for transportation to the surface.

At 12:12 p.m., atmosphere quality behind the 8 North temporary seals was taken at the end of
a sampling tube that had been installed from the seal inby for 300 feet. The quality was 20.1%
oxygen, 1,591 parts per million of carbon monoxide and 2.61% methane.

At 1:00 p.m., a new fresh air base was established at crosscut 135. At 2:01 p.m., the fresh air
base was moved to survey station 24202 in #2 entry of #22 Headgate. Temporary ventilation
controls were being installed between the #1 belt entry and the #2 track entry as the team
advanced toward the face of #22 Headgate.

At 2:59 p.m., two mine rescue teams entered the mine.

At 3:52 p.m., the fresh air base requested permission to move to crosscut 7 of #22 Headgate.
Permission was granted.

At 4:50 p.m., the team exploring #22 Headgate took an air quality measurement at crosscut 14.
The quality was 17% oxygen, 175 parts per million of carbon monoxide and 2.0% methane.
47

State command center notes 4/12/2010, page 38

13

SECTION 5

 At 5:36 p.m., a mine rescue team observed smoke at crosscut 116. There was 4 parts per
million of carbon monoxide and 0% methane. The hot spot was smothered and two team
members remained at the site.48

At 6:50 p.m., State and Federal team members walked along the curtain line to see that all
controls were installed. Rescue team members were getting low on oxygen.

At 7:25 p.m., the recovery team arrived at the #22 Headgate mantrip.

At 8:18 p.m., the first victim was moved toward the fresh air base at crosscut 129.

From 8:29 p.m. until 8:42 p.m., 76 mine rescue team members, who included State and
Federal mine rescue team members, entered the mine. Once this group of mine rescue
personnel reached the fresh air base at crosscut 78 they were directed to spread out about every
six crosscuts from that fresh air base inby to the last victim on #22 Headgate section, a
distance of approximately 8,250 feet. Team members carried each victim from four to six
crosscuts before passing them off to other team members. The walkway was often obstructed
with debris and the route of travel was not often in a straight line. At least 140 mine rescue
team members forming a human chain were used to remove the victims from the #22
Headgate section to crosscut 78 where they were placed on mantrips for transportation to the
surface.

April 13, 2010
At 2:30 a.m., a mantrip with five victims arrived on the surface at the North Portal.49 At 4
a.m., a mantrip with four victims arrived on the surface at the North Portal.50
At approximately 5:30 a.m., all mine rescue team members were on the surface. No one was
underground. Recovery was complete.
48

State command center notes 4/12/2010. Book 2, Page 8
Whitesville Vol. Fire Department Timeline
50
Whitesville Vol. Fire Department Timeline
49

14

SECTION 5

 Mine Rescue Teams that participated in the rescue and recovery of the Upper Big
Branch Mine

Alpha Natural Resources

Patriot

Brooks Run – North

Federal #2

Brooks Run – South

Southern Appalachia

Rock Spring – Gold

Magnum

Rock Spring – Blue
Kingston – White
Kingston – Red

ICG

Cobra

Wolf Run – Blue
Wolf Run – White
Hazard – Flint Ridge Blue
Knott County – White
Beckley – Black
Beckley – Gold

Black Mountain Resources

Mountaineer Mine Rescue Assoc.

Kentucky – Blue

Mountaineer 1

Kentucky – White

Mountaineer 2

Cumberland Resources

Southern Pocahontas

Virginia – Maroon

Apache

Virginia – Black

Cherokee

Arch Coal

Consol Energy/Coal River Energy

Mingo Logan – Mountain Laurel

Coal River – Red*

Gold*

Coal River – Blue*

15

SECTION 5

 Massey Energy

MSHA

Southern WV 1

Beckley

Southern WV 2

Pittsburgh

East Kentucky
Knox Creek
West Virginia Office of Miners’ Health, Safety & Training
Westover
Welch
Danville
Oak Hill

*Teams arrived at the mine on 4/5/2010 and were on standby but did not go underground.

16

SECTION 5

 5.2

List of Victims

Figure 1. General locations of the deceased miners

Carl C. Acord, 52, roof bolter operator, #22 Tailgate section
Jason M. Atkins, 25, roof bolter operator, #22 Tailgate section
Christopher L. Bell, Sr., 33, utility person, #21 Longwall section
Gregory S. Brock, 47, electrician, #22 Headgate section
Kenneth A. Chapman, 53, roof bolter operator, #22 Headgate section
Robert E. Clark, 41, continuous miner operator, #22 Tailgate section
Charles T. Davis, 51, foreman, #21 Longwall section

17

SECTION 5

 Cory T. Davis, 20, apprentice, #21 Longwall section
Michael L. Elswick, 56, underground beltman fireboss
William I. Griffith, 54, continuous miner operator, #22 Headgate section
Steven J. Harrah, 40, production foreman, #22 Tailgate section
Edward D. Jones, 50, production foreman, #22 Headgate section
Richard K. Lane, 45, production foreman, #21 Longwall section
William R. Lynch, 59, shuttle car operator, #22 Tailgate section
Joe Marcum, 57, continuous miner operator, #22 Headgate section
Ronald L. Maynor, 31, scoop operator, #22 Headgate section
Nicolas D. McCroskey, 26, electrician, #21 Longwall section
James E. Mooney, 51, shuttle car operator, #22 Headgate section
Adam K. Morgan, 21, apprentice, #21 Longwall section
Rex L. Mullins, 50, headgate operator, #21 Longwall section
Joshua S. Napper, 25, apprentice, #21 Longwall section
Howard D. Payne, Jr., 53, roof bolter operator, #22 Headgate section
Dillard E. Persinger, 32, shield operator, #21 Longwall section
Joel R. Price, 55, shearer operator, #21 Longwall section
Gary W. Quarles, 33, shearer operator, #21 Longwall section
Deward A. Scott, 58, shuttle car operator, #22 Tailgate section
Grover D. Skeens, 57, maintenance foreman, #21 Longwall section
Benny R. Willingham, 61, roof bolter operator, #22 Tailgate section
Ricky L. Workman, 50, shuttle car operator, #22 Headgate section

18

SECTION 5

 5.3

Monitoring Conditions in the Mine

Part of the underground response to mine emergencies from fires or explosions involves drilling one
or more boreholes into the mine to monitor its atmosphere to determine if the mine is burning and if
there is a risk of explosion. Secondary explosions after a mine explosion are of great concern.

Figure 2: Location of some of the boreholes

In the late hours after the explosion on April 5, 2010 preparations were made to begin drilling
boreholes of approximately 8 inches in diameter, and also to establish a seismic monitoring station to
listen for signaling from possible survivors. A mobile response truck carrying seismic equipment was

19

SECTION 5

 set up on the surface above #22 Headgate and began monitoring on April 6, 2010. The #22 Headgate
was inaccessible for underground exploration by rescue teams due to the buildup of noxious and
explosive gases after the explosion destroyed the ventilation system. The seismic monitoring detected
no activity.

The first successful borehole to penetrate the mine was drilled near the end of #22 Headgate. This
hole successfully penetrated to the Eagle seam at a depth of approximately 1,093 feet at 4:10 a.m. on
April 7, 2010. Mine voids were encountered at approximately 75 feet and 380 feet depths. Steel
casing was installed to a depth of 400 feet. This was remarkable given depth of drilling, the speed of
drilling while keeping the hole on target, and successfully drilling through the abandoned mine works
of the overlying Winifrede seam for which there were no available detailed mine maps to determine
where a solid coal pillar might exist.

Additional boreholes were drilled for monitoring, and some of these were fitted with auxiliary fans to
help ventilate the affected areas. Others were used for injection points to inject nitrogen in order to
displace combustion gases with inert gas where ventilation could not reach.

No personnel were permitted inside UBB between April 13, 2010 and June 2, 2010, due to these gases
and the potential of active “hot spots” to reignite combustible mixtures. Hot spots of glowing embers
were still found in a few locations when the mine was reentered in June 2010.

20

SECTION 5

 6

Administrative Inquiry

6.1

Regulatory Compliance

6.2

Mine Examinations

6.3

Certifications

6.4

Rock Dusting

6.5

Refuge Stations

6.6

Inspections Prior to the Explosion

 6

Administrative Inquiry

6.1

Regulatory Compliance

Violations Issued Post-explosion by the WVOMHS&T

Since April 5, 2010, WVOMHS&T personnel have conducted both an investigation of the
UBB explosion and continued required inspections at the mine. During the course of the
investigation, the WVOMHS&T issued 253 violations which include 1 Order and 2 Failure to
Abate Orders. Twenty- two of the violations were special assessed. In addition, 3 Individual
Personal Assessments (IPA) and 1 Withdrawal of Certification (WC) was issued.
Violations by category: 15 Ventilation; 11 Equipment; 14 Foreman; 12 Fireboss; 1 Haulage; 6
Transportation; 2 Coal Dust & Rock Dust; 21 Roof, Face & Ribs; 1 Fire Protection; 147
Electricity; 1 Safeguards for Mechanical Equipment; 4 Surface Structures & Practices; 1
Underground Workings; 9 Miscellaneous Safety Provisions; and 8 Other Categories.
Furthermore, after a post-explosion audit of the company’s accident history prior to April 5,
2010, 28 additional violations were issued for failing to notify the WVOMHS&T of reportable
accidents.

1

SECTION 6

 6.2

Mine Examinations

Mine Examinations and Improper Record Keeping

The mining operations of UBB are mandated to follow the mining laws of the state of West
Virginia. These requirements are available in written form, and are required to be provided to
every miner upon his or her employment at any mine within the state. The operator is to see
that all employees are properly trained in the areas of their responsibilities and necessary items
are provided them to carry out these responsibilities in a safe manner.

The law clearly places the primary responsibility for the safety of all employees on the mine
operator, in this case, Performance Coal Company. West Virginia law requires a minimum
standard be met by the operator, and only the operator can assure that these standards are
complied with on a consistent basis.

One of the concerns that arose as the investigation proceeded was the lack of specific
knowledge some of the supervisors related under oath regarding certain plans or specifics of
plans that had been put in place at the mine. This was most evident when supervisors were
questioned about the current #21 Longwall Section Approved Methane and Dust Control Plan
and the firebosses’ lack of specific knowledge regarding the requirements in this approved
plan.

During the investigation, testimony varied as to how workers performed their fireboss and
supervisory duties. It is impossible to draw solid conclusions other than that additional
training should have been given to assist firebosses and supervisors in their understanding of
the plans and the importance of following all approved plans relating to their specific areas of
responsibility.

2

SECTION 6

 Those who were responsible for the firebossing of the mine communicated by testimony that
they felt in most instances they were given adequate time to conduct their examinations.
Occasionally they would be rushed due to the need to correct problems or to take care of
concerns they had found during the course of their examinations. In some instances they felt
their other duties, including filling trickle dusters or shoveling and rock dusting belt heads,
rushed them in their firebossing efforts and sometimes prohibited them from completing their
examinations before the end of their shifts. All indications were that examiners were provided
with the essential equipment necessary to perform their jobs.

Obvious problems observed while reviewing the pre-shift and on-shift books was the lack of
clarity and full disclosure of findings, failure to accurately record corrections made and
incomplete entries on work done to clear hazards or violations noted in the examination books.
Extreme brevity of information was used on a daily basis in these record books in areas that
additional information could have been helpful to those on the oncoming shifts. Also, there
were many instances where there was no record of safety meetings or contacts required by law
with workers prior to them engaging in certain activities, i.e. roof control plan review prior to
securing the mine roof. There were omissions in the record books from one shift to the next
where it was clear that work had been conducted, but little or no information was given to
explain how conditions changed. Accurate record keeping in the fireboss books is one of the
most reliable forms of communicating to mine management and the oncoming shift the
hazards in the mine at the time examinations are performed. If used properly, these records are
designed to be a useful tool to assure that hazards and conditions found are corrected to
provide a safe workplace. State law requires the mine foreman of a mine to countersign the
fireboss books, and to make necessary steps to assure all hazards and violations are corrected
or promptly addressed. It is our belief that record books that were countersigned regularly by
those acting as “Mine Foremen” at Performance Coal Company often contained information
that should have been dealt with in a more compliant and timely manner and with more
regularity. It was obvious from our review of numerous pre-shift and on-shift books that there
was much to be desired in regards to the mine foreman’s involvement in seeing that these
books were kept and maintained in a usable fashion. This was most evident in the records of

3

SECTION 6

 the belt firebosses who regularly recorded the need for additional rock dust along the mine
conveyor belts in the explosion areas inby the switch for North Glory Mains.1

As mentioned earlier, these record books, when filled out completely and accurately, should
reflect not only the conditions found, but also the steps taken in each instance to correct the
noted hazards. Failure to reflect the actual conditions found and the steps taken to correct each
hazard and violation noted in the record books increase the risk of unnecessary exposure to
unsafe conditions and affect the correction of these conditions in a timely manner. It is our
belief that the lack of diligence in correcting the constant, reoccurring need for additional rock
dust along the belt conveyors in the areas previously referred to contributed to the propagation
of the explosion that occurred on April 5, 2010.

6.3

Certifications

Performance Coal Company provided an employee master list with all active and contract
employees that were assigned to UBB at the time of the explosion. All employee names that
were provided were checked through our safety information system (SIS) and found to have
appropriate certifications.

6.4

Rock Dusting

The WVOMHS&T has concluded that inadequate rock dusting practices and standards
contributed to the propagation of the coal dust explosion that occurred on April 5, 2010.

1

Examples of re-occurring conditions that existed in the conveyor belt pre-shift record books. (see Appendix 6.2-1).

4

SECTION 6

 One of the most disturbing facts relating to rock dusting issues at UBB was that according to
testimony the #21 Tailgate side of the longwall was never rock dusted after the longwall
started production in September 2009. The only rock dust applied in this area of the mine was
when the section was advancing on development. It is unclear why these active return airways
were neglected as part of the regular rock dusting regimen.

The outside southern entries of the tailgate were return entries for the #22 Headgate and #22
Tailgate sections and were separated by a stopping line from adjacent entries. The entries
nearest the longwall, or the most northern entries, were the return for the active longwall as it
retreated. This return pulled toward the Bandytown fan. The #21 Longwall Section did not
utilize a trickle duster or other means of mechanical dusting on the tailgate side of the longwall
while mining the 1,000 foot wide longwall face.

Approximately 5,400 feet of the 6,700 foot-long panel had been mined without any record of
rock dust being applied to the tailgate side entries. There was no testimony or records found to
dispute these findings. Witness statements indicate that the bulk duster could not travel the
tailgate side of the longwall due to low roof clearance in the track entry.

Another disturbing fact is that even though trickle rock dusters were placed near the Mother
Drive, the #22 Headgate Drive and the 6 North Glory Mains Belt Drive, firebosses who
traveled these areas on a daily basis routinely recorded in their examination books that these
areas still needed additional rock dust (see Figure 1). These examination book entries were
frequently carried over for days and even weeks at a time.

5

SECTION 6

 Figure 1. Location of “trickle dusters” found in the explosion area during the investigation.

Also, testimony, the notes provided by rock dusting crews and maintenance records indicate
that the “orange” A. L. Lee track mounted bulk or double pod rock duster used in months prior
to the explosion was quite frequently unavailable due to ongoing mechanical issues. Another
track mounted bulk duster (white) was parked in the mine yard but had not been used in
sometime due to major mechanical problems. The orange track mounted bulk rock duster was
not a dependable or reliable means to apply rock dust sufficiently on a daily basis to a mine of
this size. Testimony and written records indicate that when the bulk duster was down, rock
dusting would occasionally be done by hand. From all information available, rock dusting
procedures were not systematically organized, and equipment needed to keep a regular rock
dusting program in place was not readily available. Mine management appears to have done
little to insure that rock dusting equipment was maintained in a dependable fashion.

In addition to the concerns with the belt entries and the tailgate side of the longwall as
described above, other entries throughout the mine in outby areas that were not near the track
entries were not rock dusted on a regular and consistent basis.

6

SECTION 6

 6.5

Refuge Stations

Strata Emergency Shelter/Refuge Chamber Deployment Test March 31, 2011

A test of the required emergency shelters was conducted at UBB on March 31, 2011.
Representatives from Strata Safety Products were at the mine site to perform the test. They
were accompanied by representatives from Performance Coal Company and members of
investigative teams from the WVOMHS&T, MSHA, UMWA and the Governor’s
Independent Investigation Panel. A deployment test was conducted underground on the three
Strata shelters located in the explosion area.

Figure 2. Strata Emergency Shelter/Refuge Chamber; #22 Headgate, approximately 20 feet north of
survey station 24548 in the #2 to #3 crosscut approximately 1,060 feet outby the furthest advanced #1
face.

The first of the three shelters deployed was located on the #21 Longwall Section,
approximately 20-feet north of survey station 22738 in the #2 to #3 crosscut approximately
240 feet from the longwall working face. Thick soot and dust was impacted on the south, or
deployment door, end of the shelter. A visual examination of the outer framework of the

7

SECTION 6

 model #M2624-3.5 shelter showed there was no obvious damage to the shelter other than one
of the deployment door latch seals was partially pulled out. It also appeared that the southern
end of the shelter was moved slightly in an east to west direction by the force of the explosion.
The control panel doors were opened and the gauges and pressures were checked and found to
be in satisfactory condition with adequate pressures indicated. The pins and bolts were
removed, deployment doors were opened and the inflatable tent was pulled out approximately
four feet by hand. The “rip cord” was pulled and air cylinders activated to inflate the tent.
Although there were some small traces of soot inside the deployment door lying on the bottom
of the shelter, everything was fully operational and the shelter inflated as designed. Some
members of the teams entered the tent and assisted in getting the necessary supplies from the
storage area of the solid structure to activate the CO2 scrubber system. The system was
activated and functioned properly. The other items in the shelter storage area were in good
condition and usable at the time of the test.

The second shelter, model #M2624-3.5, was located on the #22 Headgate Section,
approximately 20-feet north of survey station 24548 in the #2 to #3 crosscut approximately
1,060 feet outby the furthest advanced #1 face. This shelter also was moved slightly by the
force of the explosion in an east to west direction on the south end. This shelter was examined
and deployed in the same fashion as the longwall shelter with the same satisfactory results.
There was no physical damage to the shelter. The gauges and air cylinders were in working
condition with pressures on tanks at appropriate levels upon inspection. The only indication
that the shelter had been through an explosion was that small traces of heat had entered the
deployment door at three of the four corners of the door.

The third shelter, model #M2624-3.5, was located on the #22 Tailgate Section, 25-feet east of
survey station 24469. The shelter is in line with the #3 entry of the #22 Tailgate Section,
approximately 850 feet outby the #3 Face. When deployed, this shelter was in working
condition. Some small traces of dust were located inside the deployment door and inside the
control panel, and some visual signs of heat had entered through the emergency door latch. In

8

SECTION 6

 addition, one of the hinge welds for the protective bar at the deployment door was broken.
None of these conditions had any impact on the deployment of the shelter.

It is encouraging that the units performed as designed after being through an event the
magnitude of the explosion at UBB. It is also worth noting that the three units deployed had
not been serviced or altered for nearly a year since the explosion yet was capable of
maintaining their usability and durability over an extended period of time without any
maintenance or upkeep.

Mine Lifeline LLC 96 Hour Breathable Air Solutions—Outby Refuge Alternatives

There were two Mine Lifeline LLC 96 Hour Outby Refuge Alternatives in the explosion area.
One was located in the #21 Longwall Tailgate entries at crosscut 9, directly north of survey
station 22145 between #3 and #4 entries.

The second outby refuge alternative was located in the North Glory Mains at crosscut 89
between #3 and #4 entries, directly northwest of survey station 19698. The forces and
pressures of the explosion destroyed the concrete block partitions and damaged the
components of both of the erected outby refuge alternatives. The solid block walls erected as
enclosure walls for the workers’ refuge area were almost completely displaced. The bottom
row of blocks and a small portion of the north end wall on the Outby Tailgate Refuge
Alternative remained in place. Metal compartments with airlock doors integrated into the prebuilt block room, where the workers would be housed during a disaster, were dislodged.
Some of the access doors provided to the compressed air cylinders were damaged and open.
Numerous compressed air cylinders enclosed in the shelter area were moved by the explosion
forces but were intact. Although these shelters are not required by West Virginia Mining Law
they were put in place as an additional alternative for workers in the event of a mine
emergency.

9

SECTION 6

 Figure 3. This Mine Lifeline LLC Outby Refuge Alternative was located in the North Glory Mains at
crosscut 89 between #3 and #4 entries, directly northwest of survey station 19698.

6.6

Inspections Prior to the Explosion

Summary of Violations for UBB (2009 and First Quarter 2010)

The WVOMHS&T issued 287 violations and five orders from January 6, 2009 to December
30, 2009. During this period, there were 137 inspection days and 17 different inspections.
These inspections consisted of four regular inspections, three investigation inspections, one
roof control inspection, eight check inspections and one electrical inspection. The electrical
inspection for 2009 totaled 21 inspection days but was not completed.

10

SECTION 6

 Violations by Category: 74 Electricity; 34 Roof, Face & Ribs; 26 Transportation; 21 General
Safety Provisions; 20 Other Categories; 17 Ventilation; 18 Foreman; 15 Equipment; 7
Safeguards for Mechanical Equipment; 6 Mine Maps; and 5 Fireboss.

In the first quarter of 2010, 39 inspection days were recorded and 44 violations were issued.
This included one regular inspection and seven check inspections from January 6, 2010 to
March 31, 2010.

Violations by Category: 11 Electricity; 7 Coal Dust & Rock Dust; 6 Roof, Face & Ribs; 4
Transportation; 3 Foreman; 3 Miscellaneous Safety Provisions; 3 General Safety Provisions; 2
Other Categories; 2 Equipment; and 1 Surface Structures and Practices.

Due to the size of this mine and its extensive outby travelways, two district mine inspectors
were assigned to UBB as of April 1, 2010, which began the second quarter of 2010.

11

SECTION 6

 7

Investigation into
The Cause of the Explosion

7.1

Structure of the Joint Investigation

7.2

Interviews and Testimony

7.3

Underground Ventilation System

7.4

Mine Seals

7.5

Sources of Ignition

7.6

Electrical Equipment

7.7

Evidence Documentation

7.8

Sampling of Mine Dust

7.9

Documentation of Flames and Forces

 7

Investigation into
The Cause of the Explosion

7.1

Structure of the Joint Investigation

On Tuesday, April 20, 2010, the initial investigative team of the WVOMHS&T was
established. A meeting was held with the MSHA investigation team on Wednesday, April 21,
to introduce the two teams, and start preparations for a joint investigation. A meeting was held
later that day at Liberty High School to introduce the teams to the families. Also attending the
family meeting was the Governor‟s Independent Investigation Panel led by Davitt McAteer.

Normally interviews do not begin until after the investigative teams have viewed the accident
site. That did not occur during this investigation due to the dangerous conditions inside UBB
that prohibited reentry into the mine by investigtors until June 29, 2010. After reviewing the
initial documents provided by Performance Coal Company, interviews began on May 10, 2010.
During this time, UBB employees, mine rescue personnel and state and federal mine inspectors
were interviewed.

The MSHA investigative team presented the underground protocol to be followed during the
investigation. WVOMHS&T requested several changes be made to give the state investigative
team rights equal to those of the federal team during the investigation. These changes were
agreed upon. The underground investigation began on June 29, 2010. The parties involved in
the underground portion of the investigation included WVOMHS&T, MSHA, Governor‟s

1

SECTION 7

 Independent Investigation Panel, Performance Coal Company, including Massey Energy and
any of its related entities aka the Company, and duly recognized representatives of the miners
of UBB, including the United Mine Workers of America (UMWA). After the investigation
was well underway, the Moreland and Moreland law firm went through the legal process with
MSHA and were given miners‟ representatives rights; however, they never traveled with
WVOMHS&T investigators underground to participate in the investigation.

The underground investigation primarily consisted of the following teams: 5 Mapping Teams,
10 Mine Dust Survey Teams, 3 Electrical Teams, 3 Photography Teams, 1 Flames and Forces
Team, 1 Geology Team and 1 Evidence Gathering Team. The number of teams varied each
day throughout the course of the investigation.

WVOMHS&T is responsible for doing its own investigation as required by West Virginia
Code 22A-2-66 (d) and 22A-2-68. There are many obstacles to overcome during a joint
investigation of such magnitude. During the course of the joint investigation, MSHA scheduled
the teams and where they would go. The WVOMHS&T realized at some point the state team
would require additional time underground to complete its investigation. While it was a joint
investigation, we still analyzed the information independently of the other entities and draw our
own conclusions. On September 13, 2010, the state Flames and Forces Team continued on
their own gathering detailed information from the explosion area.

7.2

Interviews and Testimony

As part of the joint investigation, the WVOMHS&T and MSHA interviewed 269 people. The
Governor‟s Independent Investigation Panel also participated in most of the interviews. The
interview process began on May 10, 2010, and finished on October 6, 2011. Originally all
interviews were conducted on a voluntary basis; however, due to a large number of individuals

2

SECTION 7

 failing to appear for their interviews at the scheduled times, the WVOMHS&T began issuing
subpoenas to all witnesses in August 2010.

Nineteen people subpoenaed asserted their rights under the Fifth Amendment of the United
States Constitution.1 Seventeen of the 19 asserting their rights under the Fifth Amendment
never testified. One individual who originally asserted his Fifth Amendment rights later agreed
to a voluntary interview.2 Another who originally testified asserted his Fifth Amendment rights
when subpoenaed to come for a second interview. Almost all interviews took place at the
National Mine Health and Safety Academy in Beaver, West Virginia. A few interviews were
held in the witnesses‟ homes or at the Charleston office of the WVOMHS&T.

7.3

Underground Ventilation System

Operating Fans

Ventilation was induced into the mine by three main fans. Fans installed in drift openings at the
North and South Portals were operated blowing, and the fan installed on the 16-foot diameter
shaft, called the Bandytown fan, was an exhaust fan. Some have referred to this fan as a
bleeder fan. Blowing fans induce pressure above atmospheric pressure, which is at a maximum
just inby the fan. This positive pressure causes air movement into the mine away from the fan.
An exhaust fan creates a negative pressure, which is at its maximum just before the air enters
the fan from the mine (see Table 1 for information about the air quantity of each.)

1

List attached in Appendix 7.2.1
Rick Nicolau originally asserted his Fifth Amendment rights but later voluntarily agreed to an interview on
May 19, 2011.
2

3

SECTION 7

 All values in cubic feet per minute
(cfm)
Fans, Intakes and Returns
South Portal, #3 Fan
North Portal, #2 Fan
East portal, Return
Ellis Portal, Intake
Total Intakes & Returns
Difference
Bandytown, # 4 fan
Total Intakes
Total Returns
Difference

Intake
238,800
501,400
43,000
786,200
175,100

Return
95,700
276,400
238,900
611,100
402,900

786,200
1,013,900
227,700

Table 1: Quantity of air at fans and other intakes and returns compiled from mine records. Note that
227,200 cfm more air leaves the mine than enters.

The South Portal #3 fan mostly ventilated the south area of the mine and had little influence on
the Bandytown fan, which provided the ventilation to the explosion area. The North Portal #2
fan forced air into the mine, and a portion of this air eventually exited the mine through the
Bandytown fan #4 (there wasn‟t a #1 fan). The North Portal fan ventilated the North Mains and
North Parallel Mains section, abandoned areas on the east side of North Parallel Mains and
neutral entries of Ellis Portal. Its influence continued to the Old North Mains where some slight
positive pressure remained in the air as it entered the Bandytown area of influence just inby the
junction to the Old North Mains. Also, the Bandytown fan received a small portion of its air,
maybe 43,000 cubic feet per minute (cfm), unassisted from the Ellis Portal. These two fans
operated simultaneously on the air current from the North Portal as it transferred between areas
of influence, and a small amount returned to the North Portal after ventilating the seals installed
around the large sealed area south of the Old North Mains. The Bandytown fan then provided
the ventilation pressure to ventilate the area of the explosion, which included the active
longwall and its associated gob area, both #22 Headgate and #22 Tailgate continuous mining
sections, the belt entries to these active sections, the longwall bleeder entries and the intake air
currents to these sections. The Bandytown fan also provided ventilation to all the open
workings associated with this mining area.

4

SECTION 7

 Active Section Air Courses

The active longwall tailgate had seven entries, five of them were ventilated with intake air that
had not ventilated a working face or belt entry; these were called neutral entries. Four of these
neutrals were common. The remaining neutral (#7 entry of #21 Tailgate) was separated from
the other four by ventilation stoppings and also contained intake air. The air currents of the #7
entry ordinarily diluted the longwall return air currents and moved them straight inby #7 entry
to the Bandytown exhaust fan. However on April 5, 2010, there was a roof fall in #7 entry
approximately 45 feet inby the longwall. Therefore, on this day the #7 entry air current mixed
with the longwall face return at the first open crosscut outby the face and traveled to #6 entry.
Also, the ventilation plan provided for air to travel around the last shield and inby to ventilate
that corner of the gob.

Two other tailgate entries, #1 and #2, contained return air from the continuous mining sections.
A typical excerpt from a longwall ventilation plan is shown in Figure 1. Each continuous
miner section had three entries, and three separate air currents were required for each section:
an intake separate from the conveyor belt, a ventilation air current for the belt which could not
be used at the face3 therefore was required to flow away from the face, and a return air current.
The idle section north of #22 Headgate also had three separate air currents, although that air
could be used as intake to #22 Headgate since it had not ventilated any working faces. The belt,
or neutral air, was required to flow outby away from the sections. These neutral airflows
entered the two continuous miner sections‟ combined return, most of it at an overcast outby the
longwall section where the return crossed the longwall belt.

3

Use of belt air is permitted with special precautions if approved by MSHA and WVOMHS&T, (see 30CFR
Part 75.350 and 22a-2-4a), but such use was not permitted at this mine on April 5, 2010.

5

SECTION 7

 Figure 1: Typical longwall face ventilation plan included in Ventilation Plan approved December 18,
2009, showing typical ventilation and monitoring points (MP).

Overcasts were used where air currents crossed; however, several sets of airlock doors were
used to provide access between the intake, return and neutral, or belt, air courses. For example,
when traveling on the main track entry of North Glory Mains, which is ventilated by neutral, or
belt air, one must go through airlock doors to enter the intake on the sections. See Figure 2 for
a ventilation schematic illustrating the basic ventilation layout.

6

SECTION 7

 Figure 2: Schematic diagram of the approved ventilation plan at the time of the explosion

The Airlock Doors

The primary purpose of mine ventilation is to provide a moving air stream to the working faces
to dilute and remove noxious and explosive gases, and to dilute and remove respirable dust
generated during the mining process. To accomplish this, ventilation controls, such as
stoppings, regulators, check curtains and line curtains are installed in the working section to
direct the air to the face where it is most needed. These controls cause resistance to air flow. In
order to permit access between air currents, such as the intake and neutral air splits, an airlock is
formed with two access doors on each side that are opened one at a time to maintain the
ventilation control intact. The doors are spaced far enough apart so that personnel or equipment
can fit between the doors. Opening both doors in a set of airlock doors outby a section would
provide a low resistance route for the air to take and allow it to bypass the higher resistance of
the section, taking air away from the face and causing a decrease in the continuous miner
section ventilation.

7

SECTION 7

 Testimony was consistent that constant repairs were needed on the doors along the track. The
potential for leaving doors open and not closing them once they were traveled through was
present and an ongoing issue according to numerous statements given under oath. The doors
near the working sections were manually operated, so the assurance of them being closed was
always dependent upon the individual‟s diligence and proper maintenance of the doors.

We believe the use of multiple doors to assist in controlling the ventilating current is a result of
incomplete advance planning and frequent changes to the ventilation plan. When the
unanticipated need for a ventilation control arises, installation of doors is much faster and more
economical than permanent overcast or undercast. Mine ventilation systems can be designed in
most cases without the need for doors on the haulage roads. Exhausting fans can make the
design easier. A blowing fan system requires at least one set of doors just inby the fan. The
numerous sets of doors installed along the haulage roads were an ongoing and time consuming
proposition to maintain as an important element of the ventilation system. Proper maintenance
of the doors was essential in assuring a safe environment in which to conduct daily activities
while underground. At least 12 different sets of doors were in use at UBB at the time of the
explosion on April 5, 2010.

The mine examiners‟ weekly air readings as recorded indicate that the ventilation system
provided the required amounts of ventilating air to all the continuous miner sections.
According to these records, both sections and the longwall air current exceeded the minimum
requirements (see Table 2 for a compilation of the section air readings taken by examiners).
However, the continuous miner section ventilation was dependent upon the integrity of all the
ventilation controls between the face, return and neutral entries. Some circumstances affecting
the ventilation evidently occurred because considerable conflicting information was provided in
the interview process. Witnesses indicated equipment airlock doors were frequently left open,
and there were times when changes were made to enhance ventilation on-shift on the
continuous miner sections. Citations for insufficient air, or ventilation traveling incorrectly, had
been cited.

8

SECTION 7

 The airlock doors between the continuous miner section splits had little effect on the longwall
section because it was in the direct line for the air to travel in order to get to the return. The
longwall ventilation was dependent only on maintaining the check curtains across the headgate
entries inby the longwall face and the regulators, or doors, at the mouth of the tailgate entries.
Even if the longwall belt air reversed it would then travel across the longwall face and thereby
add volume to the face (see Figure 1, ventilation of the longwall).

Ventilation Practices

The ventilation system had some conditions that detracted from its effectiveness. The longwall
headgate and tailgate air splits were connected together at the inby end of the panel through the
longwall set-up rooms, and the headgate and tailgate entries were separately regulated inby this
connection. The regulators were inby the longwall set-up entries, which remained at least
partially open and provided a connecting path between the gate entries. Therefore, each
regulator affected both the headgate and tailgate splits, leaving the amount of air in either split
depending upon the relative resistance of the air courses. This situation caused the pressure
across the longwall gob to be equivalent to the resistance on the longwall face and air course
resistance. The returns for the continuous miner sections and a portion of their belt air which
was regulated into the return were common in #1 and #2 entries of #21 Tailgate but separate
from the longwall returns, except that some continuous miner section return may have entered
#3 entry of #21 Headgate through EP-65.4 Increasing the longwall return regulation would
result in minimal increases on the continuous miner sections and reduce the flow in the
longwall section.

4

Approximately 4,000 feet east of the fan at Bandytown these air volumes combined and traveled to the fan.

9

SECTION 7

 Ventilation Planning and Modifications

Bandytown fan was placed in operation, according to the fan examination book, on August 31,
2009, and the active longwall was started shortly after that. In September 2009, the original
plan for the longwall ventilation, approved on Sept. 4, 2009, used the belt air for intake on the
longwall face. The continuous miner section on #22 Headgate (MMU 029) was started
November 30, 2009, to develop the headgate entries for the next longwall panel. In December
the ventilation in #3 entry of #21 Headgate (the current longwall headgate), which was the
return for #22 Headgate, deteriorated due to roof conditions, bottom heave and water
accumulations. Since this entry could not be traveled, it could no longer be used as a section
return. This initiated a series of changes of the intake, return and belt air of #22 Headgate. At
the time of the explosion, and as approved on January 22, 2010, the #22 Headgate return came
down #1 entry of #22 Cross-over, joined the #22 Tailgate return (which started on March 2,
2010), then went east in #4 and #5 entries of the #21 Headgate to overcasts leading across the
longwall belt to #1 entry of #21 Crossover, then to #1 and #2 entries of the longwall tailgate.
The ventilation maps show a regulator in #31 break between #3 entry of the #21 Headgate and
#22 Tailgate return; however, there were no available air readings for this evaluation point EP65.

On December 18, 2009, a plan was approved to reverse the longwall belt air so that it would not
ventilate the face. The longwall face was still being ventilated with more than the minimum
amount of air required, but the significance of this change is that it decreased the amount of air
on the longwall face and introduced a series of changes in the ventilation system. On
December 23, 2009 the company determined they could not effectively cause the belt air to
travel outby from the face, and an interim plan was approved. This plan required the belt air to
point feed at crosscut 29 and from this point travel both ways: inby to the section and outby to
the return. The WVOMHS&T rejected this plan because it allowed potentially contaminated
belt air from #22 Headgate to enter the belt air of the longwall section, which was part of the
longwall intake. Performance Coal Company had implemented this change without approval

10

SECTION 7

 from the state. A special assessed violation was issued upon inspection. This violation was
corrected and the abatement was issued on March 2, 2010.

Table 2: Record of ventilation quantities on sections, on-shift. Air quantities are in cubic feet per minute
(cfm).

The recurring changes to the mining plan and revisions to the approved ventilation plan indicate
problems in both mine planning and mine maintenance. Some of the changes were the result of
mine development scheduling or mine airways not being maintained. Planning with engineers
can reduce problems in scheduling mine development. Engineers can estimate the progress of
development sections and determine where development sections will be on any given date.
By doing such, scheduling of development units can be more effective, requiring fewer
changes. Mine ventilation construction, such as overcasts, location of air courses, power and
haulage needs can be determined and made ready on schedule.

Because development of the next longwall panel was behind schedule, the continuous mining
section had to develop #22 Cross-over and start #22 Headgate after the longwall production had
started. This required the use of the #22 Cross-over for its coal haulage, causing the section belt
to dump on the longwall belt at crosscut 29. The section returns used part of the #21 Headgate
entries. Multiple ventilation plan revisions were necessary to create a belt haulage outlet onto
#7 North belt and relocate the section intake and returns.

11

SECTION 7

 Failure to maintain the #21 Headgate entries caused the need to relocate the continuous miner
section return from #3 entry of the headgate entries, and further approved plan changes were
needed for the unplanned start-up of #22 Tailgate section. Original planning used the existing
headgate entries for the new longwall tailgate.

Another factor, referred to earlier, entered the mix of circumstances requiring ventilation plan
revisions when MSHA and WVOMHS&T apparently required the longwall belt air to be
reversed from the direction approved on the longwall start-up. The belt air could no longer be
used to ventilate the longwall face. This change was not successfully achieved on first trial,
requiring additional ventilation plan revisions.

Recordkeeping and Map Updates

The records of weekly air readings, which are useful in ventilation monitoring, were
incomplete. Routine testing for methane content at the gob edges seemed non-existent, as
evidenced by lack of procedures to examine the approaches to the gob or air leaving the gob,
even though the mine had experienced two outbursts of gas and an ignition in the gob. A
monitoring station called MP-B was set in the approved plan to measure on a weekly basis the
ventilation at a location in the tailgate near the beginning of the gob (see Figure 1 for this
location), but the weekly readings at this location were incomplete.

There were no quality or quantity air readings recorded in the fireboss books for the longwall
tailgate MP-B locations after March 17, 2010. The only information noted concerning the MPB monitoring points in the longwall preshift book in the previously noted timeframe was a
repeated notation of the statement “air to gob”, or “movement to gob.” Testimony confirmed
that the longwall foremen did not go into these tailgate entries on a regular basis. This location
did not evaluate the methane content of air coming out of the gob in the tailgate. Had such a
location been kept under surveillance the increase in methane liberation might have been
detected before the explosion. In addition, records of the required air measurements had

12

SECTION 7

 inconsistencies in the weekly examinations by various mine examiners. Discrepancies between
total intake and return air flows are shown in Table 1.
The mine map was not kept up-to-date in that some ventilation controls were not removed from
the map when they were no longer present or used. Some doors still shown in place were not
used. A map in such outdated condition could not be an effective tool in ventilation
management or decision making, and could be misleading in emergency situations when
miners‟ lives may depend on the correct interpretation of the mine ventilation system. After the
explosion most controls were destroyed and displaced. It was difficult to precisely determine
what was where because the mine map could not be used as a reliable reference.

Uncertainties about the Ventilation System

Ventilation details have been difficult to reconstruct because the explosion damage destroyed
and displaced almost all ventilation controls within the explosion area. The examination
system where weekly air measurements were made and records kept was incomplete, and some
measurements appear to be inconsistent or erroneous. The records that were available were
studied carefully to determine the actual air quantities at the time of the explosion.

The longwall was ventilated by about 55,000 to 58,000 cfm, and on April 5, 2010, at 2:40 p.m.,
the pre-shift record showed the call-out reported 56,840 cfm. The continuous miner sections
had values recorded for the last open crosscut as follows: 36,800 for #22 Tailgate and 18,554
for #22 Headgate on April 5, 2010.
The mine map is vague as to whether any return air from the continuous miner sections traveled
through an indicated regulator in the #3entry of #21 Headgate at break 31 (see Figure 3).

13

SECTION 7

 Figure 3. Regulator in #3 entry of #21 Headgate (EP 65). Adapted from a copy of the UBB wall map.

The regulator could have been closed off. This was also an evaluation point, EP-65, requiring a
weekly air reading; none was found. The headgate entries had a quantity of 50,200 cfm
recorded in the last weekly measurement at MP-36, located at crosscut 37. A problem exists in
air balancing between MP-36 and EP LW-3, located in the headgate entries inby MP-36, where
147,000 cfm was measured the week of March 30, 2010. This is a difference of about 97,000
cfm that cannot be accounted for. It is possible the EP LW-3 location is such that some air
from the tailgate entries were measured twice, which could account for the discrepancy (see
Figure 4). Also, it appears that some of the EP LW-3 readings were taken in regulators, which
introduce opportunities for inaccuracies.

14

SECTION 7

 Figure 4. Portion of mine map showing the location on EP LW-3 and tailgate measuring locations.

The #21 Tailgate entries #3 through #6 had air currents moving toward the longwall face. The
weekly reading for March 30, 2010, at #3 through #6 entries had a combined total of about
43,500 cfm. The air velocity in these entries ranged from about 59 to 114 feet per minute
(fpm). This leaves an uncomfortable situation with such low velocities, not enough to control
methane layering5 or, for that matter, to have much energy to move methane away. The #7
entry readings were not available, but using information from the MSHA model about 10,000
cfm could have been moving inby in #7 entry. We have no record of the amount of air in this
entry. The NIOSH model (WVMHS&T only has a draft of this model) shows 51,000, and the
MSHA model shows 54,200 total in all tailgate entries.

5

Kissell, F. N. Handbook for Methane Control, National Institute for Occupational Safety and Health, IC
9486, page 12.

15

SECTION 7

 After the tailgate air joins with the face air, the combined total volume would be about 111,000
cfm. The velocity in the tailgate entries inby the longwall face would be about 198 fpm, a
better velocity for moving methane accumulations.
The investigators question the accuracy of the ventilation measurements at the Bandytown fan.
Federal inspectors recorded quantities of 374,893 cfm on January 20, 2010, and 448,200 cfm
on November 3, 2009. The latest mine examiner measurements recorded before the explosion
were about 402,000 cfm.6 These measurements were taken in high velocity air currents, near or
above 1,000 fpm, and in those conditions errors caused by measuring technique or condition of
the anemometers have a significant impact on the results.

Reconstructing the Performance of the Exhaust Fan at Bandytown
During this investigation the air quantity at Bandytown fan was investigated by other means.
Information was used from an MSHA ventilation survey made September 28, 2010, when
MSHA engineers made a careful ventilation survey in the longwall headgate and tailgate entries
and captured the air quantity moving to Bandytown fan. They found 294,200 cfm. Records of
the fan pressures for this time indicated the fan was operating at 6.5 inches water gauge. This
operating point was not on any of the fan performance curves provided by the company, but it
was an operating point for the fan in its present condition and no physical changes were
reported. It is not unusual to find this situation because the operating speed of the fan may be
slightly different than the speed used to draw the performance curves, or they may be a family
of curves developed for a slightly different diameter fan and then calculated for this fan, or the
fan blades could be at a different setting than reported. In any event, that established operating
point can be used to make an intelligent estimate of the fan output at another pressure. The new
operating point was plotted using the fan pressure of 5.5 inches water gauge, the pressure on
April 5, 2010, before the explosion. At 5.5 inches water gauge the performance curves indicate
the fan output to be about 300,000 cfm. In addition to further explore this question, calculations
were made using the quantities at the other fans and portals, correcting for increase of volume
6

MSHA Inspection record available on their website: www.msha.gov

16

SECTION 7

 due to temperature and pressure changes. A pre-explosion air quantity at the Bandytown fan
was calculated to be about 307,000 cfm.
Immediately after the explosion, the Bandytown fan pressure reduced to slightly less than 3
inches water gauge. This extrapolates to about 315,000 cfm during the first few days after the
explosion, and then the fan pressure gradually began to rise, (see Figure 5 fan performance
curves with operating points).

.
Figure 5. Performance curves for Bandytown Fan. The slope of the fan curve is such that the quantity
increases about 6,000 cfm for each 1-inch drop in fan pressure.

The volume of air at Bandytown fan is a critical part of the calculation to determine the amount
of methane liberation in a 24-hour period. The air analysis from samples collected during the
MSHA inspection on January 20, 2010, averaged 0.182% methane. The fan chart for that time
period recorded a water gauge of about 4.5 inches. At that pressure a quantity of 306,000 cfm
is estimated. Using the air volume of 306,000 cfm, the methane liberation calculates to 802,000

17

SECTION 7

 cubic feet of methane in 24 hours. Adding the 55,000 cubic feet of methane in 24 hours found
in the #1 return entry of North Portal results in a total of 857,000 cubic feet of methane in 24
hours. This return ventilated the seals containing the large sealed area south of the Old North
Mains, which contains the locations of previous methane outbursts. The air quantities used by
the inspector during the last federal inspection determined that over one million cubic feet of
methane in 24 hours (1,067,510) was being liberated from the mine. These outbursts and an
ignition that occurred in 1997 in an area that is now also sealed are discussed in more detail in
Section 3.2.

Longwall Ventilation
Although several air quantity measurements are questionable and many are missing, there
seems to be no question that the longwall face was ventilated in accordance with the approved
plan.
Management of the tailgate supports and ventilation controls were not in keeping with best
ventilation practices. The #7 entry of #21 Tailgate was found caved approximately 45 feet inby
the face to the extent that a ventilating current could not turn inby in #7 entry towards the gob.
So, rather than turn inby at the last shield and move any methane from the gob at the tailgate
junction away from the face into #7 entry, the face ventilating air current was required to turn
outby for a short distance and then turn right into the partially open crosscut along with the air
current moving inby in #7 tailgate entry. This created eddy currents in that dead air space
against the fall in #7 entry. Even though some small amount of air could move over and around
the fall, the majority did not, and that could re-circulate methane (see Figure 6).
It is recognized that caving roof behind the longwall will fall across the tailgate entry and close
it off, but best practices would require sufficient supports in the tailgate entry so as to delay the
closing of the tailgate entry immediately behind the face and maintain some air opening to the
first crosscut inby the longwall face. Stoppings typically should be removed from these
crosscuts as the face advances to aid the passage of air. This would allow the ventilation to turn
the corner in an inby direction and carry methane directly away from the face.

18

SECTION 7

 Figure 6. Junction of the Longwall with the #7 entry of #21 Tailgate. Normal ventilation path is blocked
by gob.

The airflow patterns in and around gob areas are affected by the pressure across the gobs and
the permeability of the gob as well as the method of directing air to the gob. Research has
shown that flow paths around gobs are concentrated behind the shields, near the back end of the
gob and along the tailgate entry where gob permeability is the highest.7 At UBB, the
investigators believe that most of the ventilating current on the longwall face flowed in the
space under the shields‟ canopies, and most of that was over the conveyor and in the walking
space next to the conveyor. Very little airflow occurred near the area occupied by the shield
gob side hinge and even less adjacent to the gob behind the shields. It is likely that this airflow
skirted the gob and did not penetrate any appreciable distance into the gob, leaving the large
gob area as a potential reservoir of methane.
The Eagle coal seam is not highly gaseous and, as indicated by previous events, most of the
methane liberation is from the floor. The Little Eagle coal seam lies below about 10- 15 feet
(see Section 2.2). The methane liberation from the longwall face found in the immediate return
appears minimal. Samples taken during the last federal inspection from the longwall immediate
7

Yuan, L., Smith, A.C., and Brune, J. Computational Fluid Dynamics Study on the Ventilation Flow Paths in
Longwall Gobs NIOSH, Pittsburgh, Pa.

19

SECTION 7

 return showed 0.06% methane in 61,650 cfm, which is 53,266 cubic feet of methane in 24
hours. The developing sections appeared to liberate more methane. Gas in the previously
developed longwall panel has had time to migrate out of the coal. A sample in the #22 Tailgate
section immediate return was 0.13% methane in 84,020 cfm, which is 157,285 cubic feet of
methane in 24 hours. This quantity measured in the section immediate return was higher than
any other reading found and might be from both continuous miner sections.

Figure 7. Sketch showing ventilation pressures in inches water gauge around the longwall gob, taken
from the MSHA Ventilation Model. Pressure in the airstream is ventilating around the longwall gob.
Pressures are negative and becoming more negative as one gets closer to Bandytown fan.

The longwall gob gases were probably expanding and exuding from the longwall gob at the
time of the accident due to a decreasing barometric pressure. The barometric pressure had
started increasing about 9 hours before the explosion, and about 11a.m., or 4 hours before, had
begun a decrease. At the time of the explosion it had decreased 0.07 inches mercury from the
11 a.m. high. Now, 0.07 is not much of a change but that amount is equivalent to 0.95 inches
water gauge. The pressure across the gob created by the ventilating system was slightly less
than 1 inch water gauge near the face; however, inby, about mid-point in the panel, the gob
probably had no pressure across it from one side to the other (see Figure 7). Gas was likely
exuding from the cracks in the mine floor just inby the shields where methane was found during
the investigation. This gas was likely moving along the gob behind the shields and into the

20

SECTION 7

 tailgate entries inby the longwall face. It was probably present in #7 entry of #21 Tailgate
against the gob blocking the entry just inby the face.
Due to their locations in the main longwall ventilating current, the methane monitors on the
shearer and face conveyor tail drive would not detect gas behind the shields or in the severely
blocked #7 entry against the rock fall.
The examinations made by the section foreman did not detect any appreciable methane prior to
the explosion, and the methane detecting equipment on the face did not indicate a methane
accumulation before the explosion. Evidence indicates that none of the monitors or detectors
was positioned in the areas of the methane accumulations prior to the explosion. There were no
additional examinations or tests required of the mine examiners based on the previous known
gas outbursts or ignitions. There were also no specific safety precautions or practices that could
be identified with that unusual hazard that was specific to this mine.

7.4

Mine Seals

Ventilation seals were constructed to eliminate the need to maintain, ventilate and examine the
large gob areas south and southwest of the current UBB active workings; however, these seals
were ventilated with a separate split of air directed to the return. The seals are identified by
sets, and a total of 15 seal sets were installed. A total of 64 seals were installed.

A weekly examination of all seals was required. The last recorded weekly examination of seals
1-32 was March 30, 2010, and the last weekly examination of seals 33-64 was April 1, 2010.
Also required was a daily examination of one gas valve from one seal of each set of seals.
Examiners would note if the seals were “in-gassing” or “out-gassing” and the results of
methane, oxygen and carbon monoxide tests when pumping the seals. These seals were last
examined April 5, 2010.

21

SECTION 7

 A total of 12 seals were installed at Seal Sets 1 and 2. The approximate construction date for
these Micon-type seals was 2003. The Micon seals are rated at 20 psi. Seal Sets 3-5 are Strata
Packsetter-type seals. These sets have a total of 20 seals, and the approximate construction date
was 2003. The Strata Packsetter seals are rated at 20 psi. Seal Sets 6-15 are a Mitchell-Barretttype seal and have a total of 32 seals. Construction of these seals was late 2006 and early 2007.
The Mitchell-Barrett seals are rated at 20 psi.

Six seals have been replaced at UBB. Five seals at Seal Set 3 were replaced approximately
April 1, 2010. Seal 59 was replaced in Seal Set 13 approximately May 19, 2009. The seals
were replaced with Minova Main Line Tekseal-type seals, which are rated at 50 psi.
Seal Sets 14 and15 are installed outby the tailgate side of the current longwall section. These
two seal sets separate the previously mined longwall panel from the tailgate entries for the
current longwall panel being mined. Seal Set 14 has one seal, and Seal Set 15 has two seals.

Structurally, Seal Sets 8-15 survived the explosion. However, natural deterioration at the
perimeters, and/or empty water traps likely permitted some exchange of gases into the mine
after the explosion from the sealed area.

7.5

Sources of Ignition

Embers from Welding

WVOMHS&T found no evidence that embers from cutting or welding contributed to the
explosion of April 5, 2010. There were some questions raised about the use of torches and
welding on the #21 Longwall face on the midnight shift the night before the explosion.

22

SECTION 7

 Testimony clearly indicates that cutting and welding was performed on the face conveyor chain
and the headgate side cowl.
The tail side ranging arm where the “Face Side” B-lock is located had a cover missing from the
face side of this particular B-Lock. Pictures taken and observations of the area clearly show
that the cover had been cut off sometime prior to the explosion. No record was found that
would give a precise time when this particular cover was removed. Records and testimony
show that the Joy Shearer was not mining coal for nearly three hours on the dayshift of April 5
due to problems with the face side B-Lock on the tail ranging arm. The exact location where
the B-Lock was being worked on is not certain, but it is believed the work was done on the
headgate side of the longwall somewhere between shield 65 and the headgate.
As noted earlier, the owl8 maintenance crew on April 5, 2010, had been changing flights on the
face conveyor chain. This activity often involves cutting the bolts that secured the old flights
then replacing the old flights with new ones. The new bolts were then tack welded on the new
flights installed on the face conveyor chain. It is believed that this work was done near the
headgate of the longwall face.

Smoking Materials

An examination of the record books provided indicated no record of smoking articles having
been found during required searches of workers employed at the mine. Additionally, smoking
articles were not found on any of the deceased miners recovered from the mine following the
explosion on April 5, 2010.

Electrical Sparking

There is no indication that lightning or faulty electrical equipment on the longwall caused the
explosion. An in-depth inspection of all of the longwall electrical systems was conducted, with
8

Midnight.

23

SECTION 7

 attention given to damaged cables, permissibility, splices and other possible sources of ignition.
Two questionable splices in the 4,160 volt shearer cable (one located at shield 45 and one at
shield 87) were examined and found to be adequate. Two repaired places in the 4,160 volt tail
drive cable (one located at shield 91 and one at shield 105) were examined and found to be
adequate. Damage to the #6 AWG 3 conductor 110 volt lighting power supply cable at shield
171 and to the chock interface unit (CIU) at shield 171 were attributed to explosion forces
traveling from the tailgate to the headgate. The #6 AWG 3 conductor 110 volt lighting power
supply cable was damaged near shields 39 and 62. Both of these places were examined and
determined to not have been an ignition source.

Lighting power supplies, lights, communication equipment (Comtrol phones) and other
intrinsically safe components were inspected and /or tested for possible ignition sources. Both
methane monitors, one located on the shear and the other near the longwall tail, were taken into
evidence and tested and were found to be operational. No evidence of bridging out or
bypassing these methane monitors was found. The high voltage motors on the shearer, tail
drive, stage loader, crusher and head drives were tested with a 5,000 volt meggar for possible
leakage in their insulation. All of these motors passed this test. The 4,160 volt shearer cable
and the 4,160 volt tail drive cable, located on the longwall face, were also tested with a 5,000
volt meggar for possible leakage in their insulation. Both of these cables passed this test.

Welding leads were routed across the longwall face via a flat cable located in the face conveyor
cable tray. This cable was found to be disconnected from any welder located at the headgate,
thus eliminating them as possible ignition sources.

Falling Rock Ignition

It is believed that the explosion occurred behind the shields of the longwall near the tailgate,
and it is possible that it was triggered by a gas ignition in the same vicinity. As the longwall
cuts strips of coal from the face, the shields are moved up, pushing the conveyor in front of
them. Newly exposed roof then breaks behind the shields due to the overburden pressure and

24

SECTION 7

 falls, striking rock already caved or perhaps sliding along the top of a steel caving shield. As
mining was occurring and shields were being advanced, methane was being liberated behind
the shields, from the floor cracks or expansion of methane from the gob due to falling
barometer, or both, holding against the gob and migrating to the tailgate behind the shields.
The methane would not be well mixed as it moved along with the slow moving air current
along the gob, but the boundary between the methane body and air would be flammable. An
ignition source here could involve sandstone rock, which was located above the shields. It has
been demonstrated that sandstone rocks colliding and rubbing together while falling or striking
the steel of the shields can create sparks hot enough to ignite methane.9 Not all sparks created
between falling rocks or rocks striking steel are incendiary, but it is possible to ignite an
explosive mixture this way. Explosions from this mechanism have happened, even at this mine
where an ignition in the gob behind the shields at the tailgate occurred on January 4, 1997, and
was attributed to falling rock.

Evidence of heat from behind the shields is found on plastic shield and communication
components between shield 72 and the headgate. Most of these shields show heat damage,
while there is very little evidence of heat on the longwall to the south, until reaching shield 160
and from there to the tailgate. Small amounts of coking are found on shield cylinder legs
between shield 41 and shield 72. There is also coking on shield surfaces and in the back of the
shield, next to the gob, at shield 62 (Figure 8).

If methane ignited in the gob it likely burned along the gob behind the shields, where the
methane was moving to the tailgate. Where only a thin zone of explosive (5-15%) mixture
existed in the boundary between 100% gas and 100% air the gas would burn like a wick. This
wicking process is analogous to a similar process that occurs when igniting a gas burner on a
stove. The piloted flame ignition starts on one side and travels around the burner following the
flammable boundary between gas feed and the air along this flammable boundary. Similarly,
burning could occur along any methane roof layer that exists adjacent to the gob.

9

Nagy, J. Frictional Ignition of Gas During a Roof Fall, Bureau of Mines, RI 5548. Page 7

25

SECTION 7

 Methane ignited here could burn along the gob behind the shields to the tailgate where
increased turbulence and better mixing of the methane could result in the hot gas expansion and
increasing intensity as flame propagates along the methane boundary. The flame propagation
process opposite the face near or behind the shields (gob side) is similar to the January 4, 1997,
event.10 As the flame and turbulence entered the tailgate it dispersed combustible coal dust and
initiated the coal dust explosion.

B. Shield 62—Control Box (CIU)

C. Shield 63—Light cover
Figure 8. Except for the headgate and tailgate regions, coked dust inside the longwall was found only
between shield 41 and shield 72, together with other indicators showing evidence of flame and heating at
this location. (A larger map can be found in Appendix 7.9-1, Map 4).

10

In this 1997 event the flame propagated mostly between the gob and the shields (see Appendix 3.2.1).

26

SECTION 7

 Longwall Shearer Sparking

The other possibility is that a gas ignition occurred at the shearer cutting bits. Under this
scenario the methane is liberated behind the shields, from the floor cracks or expansion of
methane from the gob due to falling barometer, or both, holding against the gob and migrating
to the tailgate behind the shields. Methane then enters the area in #7 entry of #21 Tailgate
behind and around the last shield (#176). The methane begins to mix with the air current, but
out of the main air stream and away from the tail drive mounted methane monitor, and the
methane monitor on the shearer. As the recirculation occurred in this area in the #7 entry
against the gob and far rib of the entry, the methane-air mix contained non-homogenous
mixtures above and below the explosive limit. The thin boundary between these mixtures was
flammable, and a hot smear from the bits rubbing and cutting sandstone can ignite methane if it
came into contact with a flammable boundary.

The cutting bits used on the shearer had tungsten carbide tips inserted in steel (see photo,
Figure 9). The tip is very hard and abrasion-resistant, much more so than steel. If the tip
becomes worn down then softer steel of the bit can rub against the coal, or sandstone in this
case, and rubbing of steel against sandstone can leave a hot streak behind the bit. For a very
short time, about 20 milliseconds, this streak of hot material behind the bit can ignite methane.
The streak is hot enough to ignite methane for about 5 cm behind the bit; after that distance it
cools below the ignition temperature. This hot streak can be obtained with almost any bit
material but is much more likely to occur with steel rather than tungsten carbide. Research has
shown that a water spray directed to this hot streak location from a spray mounted directly
behind the bit can cool the hot streak and prevent an ignition.11

11

Courtney, W. Frictional Ignitions with Coal Mining Bits, Bureau of Mines, IC 9251

27

SECTION 7

 Figure 9: View of the tailgate shearer drum showing the cutter bits with tungsten-carbide tips. The bit
lacing is designed to maximize the interactions of the bits as they cut away the coal and /or rock. Bits
#16 and #17 are worn past most or all of their carbide tips.

The shearer was not equipped with sprays behind the bits, which is the most effective measure
to prevent frictional ignitions. At least two shearer bits located on the tail side cutting drum
showed signs of advanced wear to the bit and carbide tip. These two bits are located against the
coal face side of the tail drum, referred to as the sump side. One of the bit blocks of these two
bits showed wear of the steel shank in which that particular bit was located.

Such an ignition of poorly mixed gas and air takes some time to develop into an explosion,
sometimes up to one or two minutes.12 This explosion could propagate away from its tailgate
location as a slow moving weak methane explosion, gathering speed into the tailgate entries
where wind forces dispersed coal dust and transitioned into a coal dust explosion. A small gas
12

Nagy, J. The Explosive Hazard in Mining, Bureau of Mines, IR 1119, page 52

28

SECTION 7

 ignition of 84.5 ft3 of methane diluted to 6.5% in air is sufficient to lift coal dust and propagate a
dust explosion.13 Methane also likely burned along the gob behind the shields where the
methane was moving to the tailgate. Where only a thin zone of flammable (5-15%) mixture
existed in the boundary between 100% gas and 100% air the gas would burn like a wick. With
limited air velocity in neutral entries and more rapid influxes of methane from gob expansion
there may not be sufficient airflow to dilute and render harmless these layers.

The hot gas expansion due to unconfined burning of non-uniform gas-air mixtures produces
eddies that will grow while the flame front searches out and follows and ignites along the
flammable boundary. Some gas could have been in the tailgate entries or against the gob
adjacent to the tailgate entries, layered due to the low velocity there. This could have added to
the roof layer flame propagation, either as a continued gas explosion or hybrid gas, in
combination with coal dust in the tailgate, increasing in velocity as it moved outby, consuming
available fuel that may have been dispersed from rib and roof surfaces by local wind forces
generated by the hot gas expansion process.

Water System to the Shearer
Water is supplied to the underground areas of UBB from two tanks, each approximately
100,000 gallons, located on the mountain side above the East Mains Portal. These tanks
receive their water from three 480 volt, 150 HP pumps, located on a platform on the Coal
River, near the Load-out.

The water enters the mine in the belt entry of the East Mains Portal. Ten Rosedale Filters are
installed in this waterline, approximately 23 breaks underground. Water is then distributed
throughout the mine via a 6 inch and 8 inch plastic water line to all of the conveyor belt lines,
continuous miner sections and the longwall section.14 On the #21 Headgate Longwall Section,
the water is supplied to the Sunflo Pump through a 4 inch line. The water is filtered through a
series of Rosedale Filters, and then goes up the monorail through two water hoses (2 inches in
diameter). The water is then filtered through a wye strainer on the outby end of the gate boxes.
13
14

Nagy, J. The Explosive Hazard in Mining, Bureau of Mines, IR 1119, page 39
Danny Laverty interview (2/24/2011) Page #53 Lines 20-24, Page #54 Lines 1-16.

29

SECTION 7

 This water is then used for dust and fire suppression, fire valve outlets and running the
conveyor (face chain torque converter drive).15

Because the primary water supply (Coal River) would become muddy after excessive rains, a
separate freshwater supply from two wells provided water for the emulsion system on the
longwall.16 This was to prevent problems with the operation of the longwall shields.17

Because of these sediment issues, the drum sprays on the longwall shearer were constantly
clogged. According to statements made during interviews, to correct this condition several
drum sprays were routinely removed to allow the shearer drums to be flushed out. The shearer
would continue to operate with the water sprays removed.18 This practice was a violation of the
approved Methane and Dust Control Plan in that the required minimum water pressure of 90
psi could not be maintained.

During inspection and testing of the shearer spray system on December 20, 2010, it was
discovered that 24 of the 45 tail drum sprays were visible. A total of seven sprays were missing
from the tail drum. Water pressure and water flow tests were conducted on the shearer water
sprays.

Water Test on the #21 Headgate 7LS Longwall Shearer
Water was supplied for these tests from a 10,000 gallon tank, located on the surface at the 8A
Borehole located above the #22 Tailgate Section. The water traveled through a 7 inch metal
line, 1,260 feet down to the mine level. The water line was then reduced to a 4 inch metal line
for approximately 3,400 feet to the shearer, located on the longwall tail. A wye strainer and
pressure reducing valve were installed in this line before it was connected to a set of Rosedale
Filters plumbed with 2 inch pipe connected in a parallel configuration. A 2 inch water hose
was then connected to the water input of the shearer.

15

Danny Laverty interview (2/24/2011) Page #54 Lines 16-24, Page #55 Lines 1-16.
Danny Laverty interview (2/24/2011) Page #46 Lines 23-24, Page #47 Lines 1-11.
17
Tommy Estep interview (3/1/2011) Page #47 Lines 8-14.
18
Tommy Estep interview (3/1/2011) Exhibit #5 (Maintenance Report)
16

30

SECTION 7

 Water pressures up to 500 psi (pounds per square inch) of water were available at the shearer
for these tests. Water pressure from 50 psi to 450 psi was connected to the water inlet of the
shearer. With several tail drum sprays clogged and seven sprays missing, water pressure could
not be measured on a pressure gauge attached to a spray port on the drum. With six sprays
installed in the missing spray ports on the tail drum and 200 psi delivery pressure at 128 gpm
(gallons per minute), 95 psi was measured on the pressure gauge attached to the tail drum.

One staple-loc spray was found in the pan line between shields 59-60. This spray was buried in
the coal on the pan line.

7.6

Electrical Equipment

There is no evidence of electrical equipment malfunction or failure that can be identified as an
ignition source. Inspection of the surface electrical installations at UBB was started on May 13,
2010. An electrical inspection team, consisting of WVOMHS&T, MSHA, Performance Coal
Company and UMWA personnel inspected and/or tested the equipment. Inspections continued
on all of the surface electrical installations until the Upper Big Branch mine was deemed safe
for the investigation teams to enter. On June 29, 2010, two electrical inspection teams,
consisting of WVOMHS&T, MSHA, Performance Coal Company and UMWA personnel
began inspecting the underground electrical equipment and installations. During the next six
months, two and three electrical teams continued to inspect electrical equipment and electrical
installations underground at UBB. All of the high voltage cables from the main sub-station,
located on the surface, to Ellis Portal and up to crosscut 10 on Old North Mains were inspected
and identified. The “Green” high voltage circuit was also inspected and identified up to crosscut
47 of the Old North Mains. This circuit along with sections of the “Blue” and “Red” high

31

SECTION 7

 voltage circuits were energized to permit additional testing and to provide power for dewatering pumps.

Electrical inspections were conducted on the section electrical equipment on the #22 Headgate
and #22 Tailgate sections. Inspections were also conducted on the electrical equipment and
installations on the #21 Headgate Longwall Section. Special attention was paid to the electrical
installations and electrical equipment on the longwall section (see Section 7.5).

Outby equipment such as battery scoops, battery forklifts, de-watering pumps, battery chargers,
roof bolters and other equipment were also inspected. Various equipment components were
taken as evidence and were tested and evaluated for possible ignition sources.

A Long Airdox Shield Hauler, located outby the longwall head gate, was inspected. One set of
battery leads were burnt into; these leads were inspected and the damage was determined to
have been caused by the explosion forces.

WVOMHS&T electrical inspectors issued 168 violations during the course of this
investigation.

7.7

Evidence Documentation

Representatives from WVOMHS&T and MSHA arrived on the accident scene soon after the
explosion on April 5, 2010. A control order was issued by the WVOMHS&T to prohibit
evidence surrounding this occurrence from being disturbed until after the investigation had
been completed. The entire permitted mine area was included in the control order. MSHA also
issued their appropriate paperwork to preserve the scene of the accident. This began the

32

SECTION 7

 evidence and chain of custody procedures that were put in place and adhered to until the
investigation was completed.

Initially, all physical evidence was collected through a joint effort between WVOMHS&T and
MSHA. Many of the electrical components in mining equipment throughout the mine were
checked to rule out any possible type of malfunction with the components. Other items were
taken to be examined and analyzed to determine pressures exerted and extent of heat or flame
in the explosion area. Items such as metal and plastic signs, plastic buckets, light fixtures and
light cables, rope, various electrical cables, aerosol cans and pressurized containers were some
of the items taken for evaluation and testing.

Physical evidence was collected following the approved protocol. Photographs, written logs,
identification numbers, locations, measurements and signatures were all part of the protocol.
These protocol approvals were for MSHA to initially pick up evidence. Other parties involved
in the investigation early on had to identify items and then request those items be taken into
custody. MSHA, at some later date, would collect these marked items. After MSHA completed
their main physical evidence pick up, the WVOMHS&T evidence team continued gathering
evidence, and Performance Coal Company began their physical evidence collection.

Approximately 900 pieces of physical evidence were collected during the investigation. Many
components taken from electrical equipment in and at the mine were taken to the MSHA
Approvals and Certification Center in Triadelphia, West Virginia, for testing and evaluation.
Other electrical components were taken to different equipment manufacturers for testing, data
retrieval and evaluation. These included Joy Manufacturing Co., Matric Corp., MSA Corp.,
and SMC Corp. All parties involved in the investigation had the opportunity to attend and
observe the testing.

33

SECTION 7

 7.8

Sampling of Mine Dust

MSHA-directed Dust Band Sampling of Roof, Ribs and Floor

Ten dust collecting teams began collecting rock dust samples throughout the upper part of UBB
in June 2010. The teams were led by MSHA and accompanied by members of the
WVOMHS&T, Performance Coal Company and the UMWA. Approximately 1,803 samples
were collected. The samples were sent to MSHA‟s Mount Hope lab to be analyzed for
incombustible content and coke. They also were sent to an independent lab so the results could
be verified. As illustrated in Figure 10, areas were sampled along selected profiles outby
crosscut 67 and comprehensively sampled inby crosscut 67, the latter of which showed a
generally deficiency of compliant rock dust amounts.

Figure 10. Locations (and results) of dust band samples collected after the mine explosion

34

SECTION 7

 WVOMHS&T Sampling of “Impacted Dust” Deposits

On September 13, 2010, the WVOMHS&T began a supplemental mapping and sampling
initiative to document impacted dust deposits, which differ from the MSHA dust perimeter
samples in that they may be more representative of actual airborne dust involved in the
explosion. These are referred to in this report as “impacted dust” deposits and samples. For
discussion purposes, the types of deposits fell into one of three categories: Crevice Dust, Roof
Bolt Head Dust, and Impacted “V” Dust Cones.

1) Crevice Dust
Deposits of impacted dusts from the explosion were found in crevices of the rock ledges and
along header boards, roof bolt plates and similar structures. These are simply pockets of dust
that were captured from one direction of the explosion forces.

2) Roof Bolt Head Dust
Threaded ¾ inch roof bolt ends captured dust deposits on one side or on opposite sides, with the
thickest deposit usually found on the windward side of the initial explosion. Deposits of dust
on the heads of roof bolts were mapped and sampled across the explosion region. The most
commonly used roof bolts at UBB have a threaded tip19 which protrudes beneath a nut used to
tighten it against the roof bolt plate and the mine roof. The protruding nut and bolt captured
dust from the travelling explosion pressure wave(s), accumulating small deposits on one or two
sides of the bolt. These bolts are installed on 4-foot spacing throughout the mine so they were a
persistent sampling point. One limitation is that these samples need to be collected early in the
investigation, as humidity changes and ventilating air currents cause their rapid deterioration.
Because they were not mapped until the fall of 2010, their reliability may have been
compromised to some degree in certain places.

19

Variously called torque tension bolts, double-twist bolts, etc.

35

SECTION 7

 3) Impacted “V” Dust deposits
Cylindrical structures with diameters between 1.5 inches and 13 inches developed different dust
structures. Conspicuous, durable impacted dust deposits which formed a distinctive dust wedge
or cone running longitudinal down the long axis of the cylinder were characteristic. The halfcircumference on the reverse side was thinly dusted and the remaining regions were typically
clean (no dust). These are referred to in this report as “impacted „V‟ dust cones,” and were
observed on support jacks, belt rollers, longwall shield cylinder legs, propsetters, etc. Their
peaks point in the windward direction of the final wind forces of the explosion (see Appendix
7.9-3 for a discussion of Final Forces).

Most of the samples analyzed by WVOMHS&T were from dust deposits on roof bolts.
Approximately 362 samples of impacted dust were collected by WVOMHS&T during the
investigation. These were analyzed for coke and TIC (total incombustible content), and the
results summarized in Appendix 7.9-1.

Documentation of these dust deposits aided determination of explosion force direction, the
sequence of multiple forces, the fuel involved in the explosion and quantities of incombustible
contents of the explosion dusts. Some of this information is summarized in the maps contained
at the end of Section 7.9.

7.9

Documentation of Flames and Forces

The part of the investigation which determined the origin, direction and magnitude of the
primary explosion forces, as well as the origin and fuel involved, is referred to herein as
“flames and forces.”

36

SECTION 7

 The region of primary focus was inby crosscut 60 of the North Mains. Not all of these inby
areas were accessible particularly the regions inby crosscut 83 of #21 Tailgate and crosscut 40
of the #21 Headgate. Some of the explosion region could not be reached and mapped due to
roof falls, adverse roof conditions and/or deep water. The accessible area was large,
approximately 315 acres,20 and one walk-through of all accessible entries and crosscuts
involved a distance of just over 50 miles.21 Most areas were examined more than once and
some as many as four or five times.

Access into the explosion region was via track to about crosscut 76 of the North Mains at which
point the track forked with one fork continuing into the North Glory Mains approximately 600
feet and the other fork continuing into #21 Tailgate approximately 900 feet, but debris from the
explosion prevented further access inby by rail.22 Except for a brief time when a rubber tired
battery jeep was allowed into the #21 Tailgate to access the longwall shearer, all access and
work beyond these track terminals was by foot travel.

The Joint Investigation Team
The Joint Flames and Forces Investigation Team23 began work inside UBB on July 7, 2010,
and continued until September 12, 2010.24 A supplemental effort was then undertaken by
WVOMHS&T to develop a comprehensive Flames and Forces Map. This work began on
September 13, 2010, and continued for approximately 13 months.

20

Areal extent of the region examined, but exclusive of areas inby #21 Tailgate and #21 Headgate which
were inaccessible to investigators, for safety reasons.
21
Measured in linear footage, and comprised of the accessible mine crosscuts and mine entries
22
The active track lines inby were otherwise generally intact.
23
Comprised of MSHA, State of West Virginia (WVOMHS&T), The Governors Independent Investigative
Panel (GIIP), Performance Coal. Co., and United Mine Workers of America (UMWA)
24
A few trips underground after that date occurred as well.

37

SECTION 7

 Supplemental Mapping by WVOMHS&T

Figure 11. Index to the WVOMHS&T Flames and Forces Map. Coordinates are West Virginia State
Plane, NAD 1927, WV-S zone.

The supplemental mapping by WVOMHS&T preserved details of explosion effects and
explosion damage for the investigation and for future research. This work supplemented the
MSHA Joint Mapping Team efforts which had previously mapped debris on the mine floor, the
location of equipment and evidence of heat. The WVOMHS&T mapping effort documented
remaining items such as roof pans, mine infrastructure (such as belt structure, waterlines, etc.),
and dust/coking patterns. Remnants of ventilation stoppings were also examined in more
detail.
The West Virginia Flames and Forces Map summarize the mapped information. It is
comprised of eight (8) sheets (see Figure 11), and is located in Appendix 9 of this report.
These maps served as a working blueprint during the documentation and analysis stages of the
investigation. Smaller maps were compiled by subject matter and appear in Section7.9, and a
series of narratives found in other appendices of this report to explain their meaning and use in
more detail. The principle subject categories of our mapping are summarized below.

38

SECTION 7

 1) Path of the explosion
Information contained in the West Virginia Flames and Forces Map, various compilations of
summary maps and other information were used to determine the likely initial path of the
explosion through the mine (see Map 1). After the passage of the initial forces, subsequent
wind forces followed. These appear to be the result of a variety of factors involving reflected
pressures, cooling of combustion gases in the seconds after the explosion, and, we believe, in
some cases additional fuel consumption and propagation. A sense of the relative strength of
these secondary forces comes from the bending of roof pans, which are the criteria followed to
prepare Map 2. For more information see:

Section 7.9 Summary Maps:
Map 1. Initial Path of the Explosion
Map 2. Return Forces Sufficient to Bend Roof Pans

2) Roof Pans
Roof pans were used as indicators of the direction and sequence of explosion forces and of the
degree of relative roof heating, and were documented across the explosion region of the mine.
Roof pans are pre-shaped thin steel plates, about 1/32-inch thick, used with roof bolts to provide
additional bearing surface against the roof, and are installed between a bolt plate25 and the mine
roof. Their primary purpose is for passive supplemental roof support.

Explosion damage to roof pans was recorded using a simple nomenclature developed by
investigators. This and other map symbols document the direction of pan bending, numbers of
pans bent per unit area, sequence of pan bending (when pans exhibit multiple bends), and the
relative degree of pan rusting. This information was useful in determining the relative intensity,
direction and sequence of wind forces from the explosion, as well as some indications of
relative heating. For more information see:

25

Bolt plates are 8”x 8” square steel plates that are installed with roof bolts to provide bearing surface for the
roof bolt against the mine roof. When roof pans are used, they are installed between the 8”x 8” plate and the
mine roof.

39

SECTION 7

 Appendix 7.9-2 Using Roof Pans as Indicators of Wind Pressure and Heat
Section 7.9 Summary Maps:
Map 3. Observed Degree of Roof Pan Bending
Map 4. Observed Degree of Roof Pan Rusting
Map 5. Roof Pan Bending in the ENTRIES
Map 6. Roof Pan Bending in the CROSSCUTS
Map 7. Regions of Severe Damage to Roof Pans
Map 8. Roof Pans Bent in the Direction of First Forces
Map 9. Roof Pans Bent in the Direction of Return Forces

3) Conveyor Belt Structures
Coal is transported from the working faces and to the outside via a series of conveyor belt
systems. The belt structure in the explosion region of the mine is constructed primarily with Cchannel steel beams which are bolted together to form two continuous, parallel structural rails
onto which the roller frames and conveyor belts are attached. The belt structures were in most
places suspended from the mine roof by chains, although in a few places floor stands are used.

Damage to conveyor belts varied according to the strength of explosion forces and the gauge of
the C-channel used in their construction. The greatest damage occurred where they were
impacted broadside by explosion forces, which deflected the structural laterally and induced a
rotation to hanging belt structure that was clockwise from left hand forces when viewed along
belt axis. For the most part, severe damage was confined to crosscut intersections, and damage
quickly lessened away from the intersections where the coal pillars provided shelter from wind
forces. Discernible directions of belt structure deflections were documented in the West
Virginia Flames and Forces Map. For more information see:

Appendix 7.9-4 Using Breached and Deflected Structures to Determine Explosion Forces

Section 7.9 Summary Maps:
Map 10. Direction of Belt Structure Deflection

40

SECTION 7

 4) Ventilation Stoppings
Ventilation stoppings are used in underground mines to isolate air currents between adjacent
entries. Most of the stoppings at UBB were of concrete block construction. Prior research
shows that solid concrete block stoppings can withstand static explosion overpressures of
approximately 2 psi. When they are impacted by overpressures they create a debris field which
is a useful indicator of the direction of explosion propagation and also useful for determining
which entry(s) were on the leading edge of the propagating explosion. Debris fields were
studied to distinguish materials used in stopping installation from those left over from
construction of the stopping. Observations and related details are depicted in the West Virginia
Flames and Forces Map. For more information on ventilation stoppings, their debris fields and
interpretation, see:

Appendix 7.9-4 Using Breached and Deflected Structures to Determine Explosion Forces
Section 7.9 Summary Maps:
Map 11. Direction of Breach of Ventilation Stoppings

5) Waterlines
Deflections of water discharge lines constructed of high density polyethylene (HDPE) pipe
where documented where their deflection directions could be discerned. Some of these lines
were in service at the time of the explosion, others were not. Details of the deflection directions
attributed to explosion forces are noted on the West Virginia Flames and Forces Map. These
deflections are indicative of the direction and strength of the wind forces that moved and/or
broke them, which sometimes involved multiple events. When their deflections are opposite
the direction of stopping breach, they are indicative of wind forces which arrived subsequent to
the initial forces, and are useful in assessing their relative strength and direction. Similar to the
conveyor belt structures, the strongest damage to the waterlines is in crosscut intersections. For
more information on damage to discharge waterlines see:

Appendix 7.9-4 Using Breached and Deflected Structures to Determine Explosion Forces

41

SECTION 7

 Section 7.9 Summary Maps:
Map 12. Direction of Waterline Deflection

6) Debris Transport
Most of the airlock doors and some of the electrical power centers, capacitor boxes, and other
mine structures which were transported by explosion forces or had parts transported during the
explosion are shown on the West Virginia Flames and Forces Map. This information assisted
in the determination of wind forces direction and sequence of explosion forces. Two examples
are given in Appendix 7.9-4.

7) Evidence of Extended Flame Duration
Carbonaceous laminations on the mine roof and mine dusts on the coal ribs that were
conspicuously coked and visible to the naked eye are referred to as “macro-coking”. These
deposits were observed on the mine roof in the form of cintered agglomerations and on the coal
ribs as granular deposits, which in extreme cases resembled fine popcorn. Deposits of macrocoke are indicative of slow flame or flame of extended duration. Macro-coking was found
throughout the explosion region of the mine, which gives evidence that coal dust participated
extensively as a fuel for the explosion. A map summarizing the observed macro-coking is in:

Section 7.9 Summary Maps:
Map 13. Observed Occurrences of Macro-coking

8) Impacted dust
Deposits of impacted dusts from the explosion were systematically mapped across the
explosion region of UBB using a series of mapping symbols developed by UBB investigators.
Deposits on protruding roof bolts, standing roof supports, conveyor belt structure, rock
crevices, etc., were mapped to help determine the direction and sequence of wind forces for the
investigation and for future reference.

42

SECTION 7

 Impacted dust samples that were collected and analyzed for coking content and total
incombustible content (TIC). They were collected primarily from the protruding threaded
heads of roof bolts. Because these deposits were created from the dust of the propagating
explosion they are useful in determining the composition of that dust, not what settled to the
floor and on ledges after the explosion.

The analyses of these impacted dust samples are in good agreement with the MSHA dust band
sample TIC analyses. They are less in agreement with the MSHA dust band sample coking
analyses, and it is possible that the difference is related to transport of dust over the mine floor
by moving air currents subsequent to the explosion. Local settlement of these dusts appears
influenced by ventilation immediately following the explosion. The impacted dust samples
collected by investigators were also studied under a scanning electron microscope for
cenospheres26, which give evidence of high temperature coal combustion. Approximately 65
samples provided to NIOSH for analysis using a scanning electron microscope showed
cenospheres present in varying quantities in nearly all samples (see Appendix 7.9-1). This is
one of the indicators that coal dust was a fuel for the explosion. For more information relating
to documentation of impacted dust, see:

Appendix 7.9-1 Sampling and Analysis of Mine Dusts
Appendix 7.9-3 Identifying Final Forces
Section 7.9 Summary Maps:
Map 14: Direction of Final Forces
Map 15: The Impacted “V” Dust Cones on Standing Supports
Map 16: Impacted Dust on Exposed Heads of Roof Bolts in the ENTRIES
Map 17: Total Incombustible Content (TIC) of Impacted Dust Samples Collected
and Analyzed by WVOMHS&T
Map 18: Coking Results of Impacted Dust Samples Collected and Analyzed by
WVOMHS&T

26

Cenospheres are microscopic glass spheres (10-50 microns), which are hollow and are indicative of coal
particles burned at high temperature at power plants.

43

SECTION 7

 Map 19: Analysis for High Temperature Glass Spheres (Cenospheres) in
Impacted Dust

9) MSHA Dust Band Samples
For reference, maps illustrating the results of MSHA dust band samples are included. They can
be found at:

Appendix 7.9-1 Sampling and Analysis of Mine Dusts
Section 7.9 Summary Maps:
Map 20: Total Incombustible Content (TIC) of Dust Band Samples Analyzed by
MSHA
Map 21: Coking Results of Dust Band Samples Analyzed by MSHA

10) Miscellaneous
Some locations of conspicuous floor heave are noted on the West Virginia Flames and Forces
Map, as were other miscellaneous items whose purposes ranged from possible heat/pressure
indicators to landmark reference points. Other items shown on the map are simply points of
reference for investigators as landmarks to confirm map locations.

Limitations of Mapping
In order to avoid duplication where possible, the West Virginia Flames and Forces Map
contains fewer details of floor debris and heat items, and fewer measured locations than the
MSHA maps. In most instances the latter should be used for more precise location purposes.
While some items in the WVOMHS&T map are drawn to scale, most items are drawn larger
for clarity. Also, with a few exceptions the WVOMHS&T Flames and Forces Map makes
fewer mapping adjustments to pillar and rib locations in the UBB base mine map than do the
MSHA maps, which results in some minor difference in the depiction of mine works.

A digital map of the UBB mine workings was provided to WVOMHS&T by UBB shortly after
April 5, 2010, and this served as the base map for our documentation and mapping effort. A

44

SECTION 7

 raster map was prepared from the digital base map and certain layers were turned off digitally
to eliminate clutter and provide space for our notations. Two legends are shown on the West
Virginia Flames and Forces Map to help identify and distinguish the data added by
investigators from the original base map features. Features shown originally on the UBB mine
map were generally not removed or relocated, even if subsequent work indicated locations on
the map were not current. Instead, our map notations often indicate approximate revised
locations to certain mine features, including doors, stoppings, regulators, etc. Also, some
survey spad numbers on the map did not match the numbers stamped on found monuments,
which for the most part appeared to be a clerical transposition.

Indicators of explosion pressure

Explosion damage to roof pans, aerosol spray containers, light bulbs and displaced structural
members were modeled by various means to estimate the explosion pressures involved. The
work was performed by State and Federal investigators. See Appendix 7.9-6 for more
information.

45

SECTION 7

 9-NORTH INITIAL PATH OF EXPLOSION

(WEST MAINS)

   

  

F. 0 500 FT.

Path of propagation

 -NORTH Propagation subsides or going to
 MAINS) ext'nctlon-

 . 3 ?low 
Om: . . 




SCALE

?09
=9 Fine interpretation

  
 
  

 

       
  

LON GWALL
HEADGATE

 
    
   
 

LONGWALL
TAILGATE

  
   
 

   
 

   

 

21 TAILGATE
(1 NORTH WLE36TELT 

 

 

SECTION 7.9 MAP 1

 

tho-I

RETURN FORCES SUFFICIENT TO BEND ROOF PANS

MAINS)

ew??b  ?g
0 500 FT. I=tsi e;  3 Path of propagation
i
SCALE 41"
ii

. ?t . . Propagation too weak to bend pans--
#22 direction indicated with Final Forces.

#22 HEADGATE "?13


. - 4:321 HEADGATE

1i  5

Fine interpretation

    
  
 
  
  
 
 
 
  
   

., - V- -. . .
LONGWALL
HEADGATE

LONGWALL
TAILGATE

PREVIOUSLY MINED BY LONGWALL AREA) PARALLEL MAINS-

 
   

  

.  I ?1 NORTH 

TAILGATE 
#21 CROSS-OVER I (1 NORTH 

 

#21 TAILGATEIR: 

'21:

 

ramoz?mamhu
-.

SECTION 7.9 MAP 2

 

OBSERVED DEGREE OF ROOF PAN BENDING

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Severe bending of Reef Pane
Fr I 
SCALE .II.
- Moderate bending of Roof Pans
- Reef Pane net bent
NOTE: Areas not shaded either have no roof pane, were
inaccessible, or were inadvertently omittedmost cases roof pane bent 90 deg. are in this
. . .. i? categorySECTION 7.9 MAP 3

OBSERVED DEGREE OF ROOF PAN RUSTING

- Roof pans mostly rusted
- Edges of roof pans rusted

Roof pans not rusted

0 500 FT.

SCALE

 
  
  
    

 

 

 

 

NOTE: Areas not shaded either have no roof pans, were
inaccessible. or were inadvertently omitted.

 

 

 

 

 

 

 

 EE-
El-v. r: i 

SECTION 7.9 MAP 4

ROOF PAN-BENDING in direction of the ENTRIES

- Roof Pans are bent to the INBY

Roof Pane are bent to the OUTBY

   
   
  
   
    

500 FT.

I_l
SCALE Black arrows- direction of pan bending

(when there is only one direction)

 

Roof Pans are bent in TWO directions

 

 

 

Red arrowe- direction of 1st Bending

Blue arrowe- direction of Final Bending

 

 

 

. . RoofPanenotbent

 

 

.u x- 44 .1 

- 
SECTION 7.9 MAP5

 

 
 

ROOF PAN-BENDING in the CROSS-CUTS

- Roof Pans are bent LEFT (looking inby)

Roof Pans are bent RIGHT {looking inby)

 
 
 
  
  
 
  
  
  
   

 

EDOFT.

SCALE

Black arrows- direction of pan bending
(when there is only one direction)

 

Roof Fans are bent in TWO directions

 

 

 

 

 

 

 

 

 

 

 

Red arrowe- direction of 1st Bending

 

 

 

 

 

 

 

 

Blue arrows- direction of Final Bending

 

 

_1l

 

 

Roof Pane are not bent

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.J 1 1 i.   13 13?ILJL


FT
43
32
L:

 

 

 

 

 

 

 

 

SECTION  MAP 6

REGIONS OF SEVERE DAMAGE TO ROOF PANS

 

 

 

 


I a 500 FT. Regions of Severe Bending" of roof
Iml pans, used as an indicator of either

Reflected Pressure (pans draped) or
High Velocity Pressure (pans uni-
directional).

NOTE: The equivalent of 1D or more pans per line
of roof pans between adjacent intersections
that have been bent 90 degrees by wind

4: d: sell f. forcesmlI-w In

SECTION 7.9 MAP 7

 Ibo I

Roof Pans Bent in Direction of First Forces

 

 

 

 

 

 

 


so. ?uni-:33
noses-om oo~"
. was,  10 roof pans bent
euro-:7 
0  6 - 9 roof pans bent
SCALE 4? 
no
I BIL-3.. -
3 5 roof pans bent
{cog-goo

0'  1 - 2 roof pans bent
 .0
 No roof pans bent
g) 8
.3, I
41? 4?  First Forces are strong enough to bend
0536-10 a? .5: Fla.
 . a a  First Forces are too weak to bend roof


pans-- direction is inferred.
 (Cl 

 

NOTE:
Pan counts are based on 1 equivalent
line of roof pans, between adjacent C,
intersections. 2 ,1 t: 
a 62;, .
I. I. cans-Ii: 0 
 no 'mniz .. 53.00.90  
no.0: no . no C5 
?lm} i  61 3-6. 
c, 0:99.332TRAIN Nu


a mo no Sou

SECTION 7.9 MAP 8

Roof Pans Bent in Direction of Return Forces

 

 

 

 

 

 

 

 

 


. 
II I 0 - 10 roof pans bent
soon  - 6 9 
- I I .13 .95, - ['00 ans 3"
SCALE ?b 
3 - 5 roof pans bent
1 - 2 roof pans bent
No roof pans bent
?p Return Forces are strong enough to bend
I =3 ES: 
  .   roof pans
Return Forces are too weak to bend roof
{5-4 ,3 pans-- direction is inferred from light-
   . weight Fina! Forces indicators, such
 c2: 61:, as paper, plastic, cloth, etc.
:3 
0.3. G)

NOTE: . 
Pan counts are based on 1 equivalent  .
line of roof pans, between adjacent a I a 
intersections. c,

1-. i



SECTION 7.9 MAP 9

Direction of Belt Structure Deflection

Black arrows- direction of Belt Structure
de?ec?on

  
  
 
 
 
 
 
 
 
  



500 FT.

 

l?I Deflection direction is inconclusive
SCALE

 

 

 

 

No signi?cant de?ection

 

NOTE: Other shaded colors are used to group de?ection
directions.

Areas not shaded either have no belt structure or they were
not mapped.

 

 

SECTION  MAP 10

 

Direction of Breach of Ventilation Stoppings

   
 
  
 
 
  

Black arrows- direction of stopping breach

. (when there is only one directionSCALE Breach direction Is Indeterminate

 

 

 

 

Stopping not breached by explosion

 

NOTE: Other shaded colors are used to group breech
directions.

Areas not shaded either have no stoppings or the breach
direction is uncertain.

. 

Stoppings across

          
 
  
  

  
 
 

     

?4 entries may not be consistent
 .. 1 with legend color assignments.
In 5; . fT? In some cases overcasts are used.
?1 :3 
I 
  ?1

  

. oil?I


  
 

 

 


  
"e-igisld?ltjeortheE-Qr -  
     .   
1- ~32? 

SECTION MAP 11

 ERNIE

Direction of Waterline Deflection

- . . Black arrows- direction of waterline

. . . - - I deflection

 

soon -- -- - .-
.  - Deflection direction is inconclusive

SCALE

 

 

No signi?cant deflection

 

 

 

NOTE: Other shaded colors are used to group deflection
directions.
4:  I Areas not shaded either have no waterlines or they were
not mappedhair?H.TRAINING
MINFH5


A we :Oc no 5m
..

SECTION 7.9 MAP 12

 

MINEHS

l?l
SCALE

  vii  

.
500 FT. .- it. I




 

SECTION 7.9 MAP 13

Observed Occurrences of Macro-coking

 

. Macro-coking of fossilfcarbonaceous
layers (approx. 2 mm) on mine roof

. Macro?coking of rib dust

Evidence of burning ?hot-spots?
at the tee of the mine ribs

 

 

 

)Iuclzmunuam

 500 FT.

SCALE

    

TRAINING
us

Direction of Final Forces



Using lightweight articles (cloth, paper,
plastic, sheet metal, etc.) as indicators.
NOTE: Lightweight debris loosely hung that would
have been easily detached by subseqent wind forces
1  coming from a different directionSECTION 7.9 MAP 14

THE IMPACTED DUST CONES ON STANDING STRUCTURES  i 

 

 

 

 

 

Onodata Bending of Roof Pans (also LW shield signs)
. TRAINING
. 

it .. 500 ?h Primary wind forces direction
SCALE - .n . . .
Onodata . Onodm Initial forces (when more than one
 direction is indicated with double-bent pans}
Subsequent wind forces (when more than one
direction is indicated with double-bent pans)
1  Lightweight Debris
 "Wm I Direction of Final Forces (as indicated by the
. -  ?nal resting position of light paper, plastic, cloth)
F?dam F25. Impacted Dust
- . .
ear.-  I Impacted dust cones on standing cylindrical
- structures (vertical pipes, jacks. cable bolt heads.
EXAMP?ggmpaggm'E DUST 0 propsetters. shield leg jacks, etc.)
a 1, NOTE: With few exceptions the dust
DUST CONE \w  0? data cone is oriented in a direction that points
l\ . 0 no data ,0 0 into the Final Force wind direction.
SURFACE - 
- -. .

- 3.
P's? 0? s: 
I V)  
EXAMPLE OF PAPER 1 74 
LOOSELY CAUGHT ON BELT STRUCTURE .. - ?0 mm I
?TLongwall Face  7;
. I .. F1 nadamok\ I. i
dug-F ?11 a 
. --ISECTION 7.9 MAP 15

 

Impacted Dust on exposed heads of Roof Bolts in the ENTRIESM

 

 

 

 

 

 

 

 

0
H, - Thickest Impacted Dust is on
6 OUTBY side of Roof Bolt
E: - Thickest impacted Dust is on
.. 500 FT. 30a: .- INBY side of Roof Bolt
Im' coco. Equal Impacted Dust on BOTH
sides of Roof Bolt

on . No Impacted Dust on Roof Bolt

6  Direction of wind forces depositing
 i the thickest Impacted Dust on Roof Bolt
0
40> 0 Symbol represents Impacted [lust on RE
on (The darkpeakls the Slde haV'ng the
. 'e thickest accumulation of Impacted Dust)
Ive-oe- concatenate-occur. 
I .l . NOTE: Because these dust geometries were not mapped until
0 .I the fall of 2010. their reliedailityr may be compromised in certain
placesExample 0f impacted nu ecu-nonun- 
Dust on Roof Bolt coo-coo 
4? .0. .
<0 o.
(21Roofbolt (11> 1? 
. . . Gilli-.0.? 'g
- . no. II one our}: 0. 
no In I on coco-co on not 0 ..
Rocfbolt I o: Iillil a
thread 3 :zigg-u. - 
3' 3' I 
(21> 03> 43>-

 

SECTION 7.9 MAP 16

 

Total Incombustible Content (TIC) of Impacted Dust Samples Collected and Analyzed by 



 

0

EODFI


SCALE



 
 

Mu.

SECTION 7.9 MAP 17

Total lncombustible Content

0 20-30%
. 304m
. 40-50%
50-60%
60-?0%


Coking Results of Impacted Dust Samples collected and analyzed by 

I 'z I a . Relative amounts of coking, based
.0 . . I on the MSHA alcohol coke test
Extra Large
SCALE
. Large
Small
(0 Trace
 None
.0,
99% 
22TRAINING
Ml'll'F?S



SECTION 7.9 MAP 18

ANALYSIS FOR HIGH TEMPERATURE GLASS SPHERES (CENOSPHERES)

 

 

 

 


I
. . Location of Samples
. analyzed by
Scanning Electron
Microscope (SEMSECTION 7.9 MAP 19

MAP 20

SECTION 7.9

 

 

 

:ii?iAQ
Iii? 1
   

qij mi:  

II 1


gnu-m  

. 

GulfQ?t??l i! :0 0 0 
ongibtt 



00006-1
qu?iobit .1403." 

. .

1P:
-


90-10080-90%

mtg-Mil? -

50-70%
70-80%

50-00%

. 40-50%

. 30-40%

. 20-30%


SCALE

500 FT.





   

pp. 


Ell.- 


Total lncombustible Content (TIC) of Dust Band Samples analyzed by MSHA
I
i 
I
II

Total lncombustible Content

 

Coking Results of the Dust Band Samples analyzed by MSHA

it?. in,? . Relative amounts of coking, based

  

.QI
"fl "Jum- lift it? on the MSHA alcohol coke test

500 FT. '1 toot-31L 

oi'i?ogiit', . Extra Large
 0 Large
Small
i:   Trace

. None

  
   
 

. L-u?sa-n mus  ..-.  . . 
 _v g: 110;;le .  

I 
irn?? "(it 


  1.-


   

I 51- - :r o?W?tf? - .-- 9..
- 
3

  

gift-S;



room-
a?SECTION 7.9 MAP 21

Source of Fuel for the Explosion

Mine Gases

Coal Dust

8
8.1

Source of Fuel for the Explosion

Mine Gases

Methane as Fuel for the Explosion

The ventilation system is central to a determination of the source of methane. A positive airflow
was moving around the perimeter of the longwall gob, which is good. However, pressure drops
across the gob between gate roads are believed to be minimal1 and, as a result, it is believed the
gob could accumulate and hold gas, especially when you consider that the gob was likely the
most gas-producing area of the mine. Evidence obtained during the investigation indicates that
the mine floor on the longwall face was a major source of methane. Air samples taken July 23
and 26, 2010, and October 26 and 28, 2010, from a floor crack near shield 160 contained
methane, even though it was some months after the explosion. Air samples near the floor on
#22 Headgate also contained methane. Analyses showed small concentrations of other
hydrocarbon gases such as ethane, propane and even hydrogen. Samples from the cracks behind
the shields when reduced to an air free analysis contained amounts of ethane and other higher
hydrocarbons similar to samples taken from a gas well. See Table 1 for the results of these air
analyses. Gases associated with coal beds typically contain mostly methane, but it is not

1

MSHA Ventilation Model

1

SECTION 8

 unusual to see other gases in small amounts as found at UBB.2 The amounts of other gases did
not significantly affect the lower explosive limit, which was calculated to be about 4.5% to 4.6%
in air for the samples collected after the explosion.

Table 1. Analyses of gas samples taken in floor cracks behind the longwall shields and on development
sections

Previous outbursts of gas occurring from cracks in the mine floor near longwall shields on
February 18, 2004, and on July 3, 2003 (see Section 3.2), provide more evidence to indicate that
the breaks in the mine floor can be a major source of gas.3 The Eagle coalbed is not very gassy,
but the longwall gob area is known to liberate gas. Further evidence of this fact is illustrated by
the occurrence of a methane gas ignition at or near the tailgate in the gob directly behind the
shields on January 4, 1997.4 This ignition did not result in serious injuries and was attributed to
friction between falling rock and/or rock striking the shields. This ignition occurred as the
shearer cut out on the tail. The shearer was cutting down and the bits were not against the mine
roof. The gas burned behind the shields on the face side of the gob for some period of time. In
2

Kim, A. G., The Composition of Coalbed Gas, Bureau of Mines, RI 7762. Kim, A. G., Experimental
Studies on the Origin and Accumulation of Coalbed Gas, Bureau of Mines. RI 8317. Kim, A. G., Gases
Desorbed From Five Coals of Low Gas Content, Bureau of Mines, RI 7768.
3
MSHA memorandum reports from Technical Support to the District Manager dated March 4, 2004 and
July 13, 2004.
4
Report of Methane Gas Ignition, Office of Miners’ Health, Safety and Training, January 4, 1997 (see
Appendix 3.2-1).

2

SECTION 8

 summary, it was known that the mine floor liberated gas, and sometimes quite a lot, as learned
from the previous methane outbursts. The Little Eagle coalbed was below about 10 to 15 feet,
and is considered to be a source of methane. See Appendix 2.2-1 for more information.

Air currents follow the path of least resistance, and air currents flowed around the gob and to the
fan following the open air courses. Little pressure incentive existed to cause airflow through the
gob. Most of the air current on the longwall face stayed in the larger area under the shields, as
evidenced by the air velocity readings taken on the longwall face. Methane that migrated out of
the floor cracks behind the shields would be induced by slow air velocity at that location to
move along the gob behind the shields to the tailgate. It most likely did not mix much with the
air current on the face and stayed relatively intact as a gas cloud until it reached the tailgate.5
Most of the air on the face under the shields made an abrupt turn to the left, outby, in #7entry,
and then after a few feet turned right through the partially open crosscut to #6 entry. It was
probably joined by air currents moving inby in # 7entry, which also turned through that crosscut.
Gob had fallen across the #7 entry inby the shields blocking #7 entry allowing only a very small
amount of the face air to penetrate that fallen area. This abrupt turn to the left resulted in part of
the air current striking the far rib, resulting in eddy currents moving in somewhat a circle
causing air recirculation in that space against the gob in #7 entry (see Figure 1 and Appendix
8.1-1). This recirculation resembles that which occurs when a river bends and creates a
recirculation pattern on the outside of the bend. The result is that methane can accumulate to
above the explosive limit in a poorly mixed zone of methane and air, and this methane would
have a direct connection to the gases moving along the gob’s edge. The zone of poorly mixed
gas and air, having complex air turbulence, could result in some explosive mixtures in the
vicinity of the shearer cutting bits as the shearer cut through the sandstone at the tailgate.

We have studied the theory that the explosion was caused when a massive inundation of gas in
the tailgate entry at the longwall face permeated the area and was ignited. Methane is light and
easily moved by air currents such as those moving across the longwall face. It is highly unlikely
that gas would have moved from the longwall tailgate against this face ventilating current and
5

Kissell, F. N. Handbook for Methane Control, National Institute for Occupational Safety and Health, IC
9486, page 12.

3

SECTION 8

 contaminated the face area into the headgate. The air in the longwall face had a velocity of at
least 700 feet per minute. Methane would not move upstream in that air current, nor does the
evidence along the face indicate that a methane deflagration occurred on the longwall face,
which surely would have occurred if flammable methane-air mixture had migrated upstream
along the longwall face.

Figure 1. Junction of the Longwall with the #7 entry of #21 Tailgate. Normal ventilation path is blocked
by gob.

The return from the longwall face combined with air in #7 entry totals about 67,000 cubic feet
per minute. Air quantities used in this discussion are from the most recent examinations when
available and when not contradicted with more reliable information. Using this quantity and the
known area of the regulator, the velocity in that regulator to #6 entry would be about 1,600 feet
per minute, sufficient to promote mixing of a body of gas from the tailgate area. A quantity of
methane exceeding 3,300 cfm would be necessary to result in an explosive mixture past this
regulator, assuming a complete mixing. Once through the regulator, the low pressure area, and
air movement were toward the Bandytown fan. Methane diffused in the air current as it flowed
through the high-velocity at the regulator into #6 entry, would not re-stratify or separate from the
air and therefore it would be carried with the air as it flowed to the low pressure area on its way
to Bandytown.

4

SECTION 8

 The air volume in tailgate entries 3 through 7 had air currents moving inby towards the longwall
and on to the Bandytown fan. A volume of about 54,000 cfm total flowed in these entries and
the air velocity ranged from about 59 to 114 feet per minute. If methane did migrate into this
area from inby it could accumulate as a layer near the roof since such low velocities are not
enough to control methane layering,6 or, for that matter, to have much energy to move methane
away. These aircourses were essentially flat, about 0.7% grade, so methane did not have much
tendency to drift outby along the roof. Methane accumulations that might have layered in the
tailgate entries would have been confined to that area because the high-velocity in the regulators
upstream would have prohibited methane from moving out of the tailgate entries.

After the tailgate split mixes with the face air the combined total volume would be at least
111,000 cfm. The average velocity in the tailgate entries inby the longwall face would be about
198 feet per minute, a better velocity for moving methane to the Bandytown fan. The leading
edge of such a methane-air body would reach the Bandytown fan in about one-half hour (the
velocity speeds up when the airways are reduced to 4 entries inby the longwall panel).
Whether an increase in methane liberation along the face was of sufficient intensity and duration
to cause some amounts of methane in the tailgate entries is unknown. The foreman in charge of
the work behind the longwall traveled through that area about 1 p.m. or 2 p.m. on April 5, 2010.
He reported the ventilation conditions as being normal but, based on the results of the download
of the information stored on his detector, it was determined that the foreman’s multi gas detector
had not been turned on at any time during his normal shift while performing work behind the
longwall on April 5, 2010.

Flame from an ignited methane accumulation along the longwall gob behind the shields or an
ignited methane accumulation of a body of gas at the shearer would certainly have traversed
some distance out the tailgate; however, it is not believed that the flame from a methane

6

Kissell, F. N. Handbook for Methane Control, National Institute for Occupational Safety and Health, IC
9486, page 12.

5

SECTION 8

 explosion at this location could have continued throughout the explosion area since methane
would not have spread outby prior to ignition.

Those with the opinion that a sudden massive outburst occurred look for evidence in the amount
of gas found after the explosion in samples at the Bandytown fan in excess of the normal mine
liberation and the relationship between ethane and methane in the samples collected at the fan.
The first analysis available was from a sample taken at 8:30 p.m. on the day of the explosion,
and it contained 0.42% methane. A WVOMHS&T inspector testified that he used his detector
in the fan discharge at 5:20 p.m. and found 2.3% methane; a repeat test some time later indicated
no methane. Without other substantive data, this information is inconclusive. No other air
quality information is available in the Bandytown fan record until MSHA collected the air
samples beginning at 8:30 p.m.

Rather than a dramatic outburst, it is more reasonable to believe that methane was liberated
along the face in increased amounts caused by the breaks in the floor and from the gob, and that
this increased liberation caused an undetected buildup against the gob behind the shields. Gas
expansion, because of the decrease in the barometer 0.07 inches mercury or about 0.95 inches
water gauge since 11 a.m., would cause additional methane to exude from the longwall gob, and
this gas could come out near the face due to the ineffective pressure drop on the gob (see Figure
2). An unusual coincidence was discovered during a review of the ignition on the gob January
4, 1997. The barometer change was quite similar to the barometer trend during this explosion
(see Figure 6).

Some methane likely entered the tailgate entries through the crosscuts to the gob just inby the
face and contaminated that area inby the face against the gob fall in #7 entry. A portion of the
face ventilation likely gradually passed through the shields to the gob along the 1,000 foot-wide
face. This leakage was not enough to dilute the gob gas to below the explosive limit, but that
flow direction, plus the amount of air remaining on the face, would have been sufficient to keep
the active face area clear. Neither the methane monitors on the face equipment nor would the
foreman with his detector on the face have detected this methane build-up.

6

SECTION 8

 Figure 2: Showing Barometric pressure decline prior to explosion

Methane was still being liberated and found during the rescue effort. At 12:24 a.m., April 6,
2010, the rescue team advanced to shield 120, their farthest advance at that time, and detected
over 2% methane. The ventilation on the longwall face would have been greatly reduced at this
time because the ventilation controls to direct air to the face had been destroyed. No mention
was made about sounds such as would be found from an on-going methane outburst.

A substantial source of methane after the explosion was the methane generated along with other
combustion products during the explosion. This methane would have been distributed
throughout the explosion area.7 The products of combustion, including the methane, were
gradually moved to the Bandytown fan by the ventilation currents.

7

Conti, D. S. et al Rapid Sampling of Products During Coal Mine Explosion, NIOSH, Pittsburgh Pa

7

SECTION 8

 Figure 3: Methane trend after the sampling started projected back to 15:30

It is believed the combustion products started arriving at Bandytown about 3:30 p.m. and
continued to be diluted and moved to the fan by the air currents. The combustion gases on the
longwall gate entries and those mine openings in direct line with them should have been purged
of the combustion gases within a few hours. Analysis of the gas trends suggest that about onethird of the combustion products had passed through the Bandytown fan by the time MSHA
sampling was started at 8:30 p.m. (see Figure 3). Remnants of combustion gases remained in
the continuous mining sections, the North area and Glory Hole area for some time until they
were flushed out during recovery. The characteristics of this gas are found at Borehole #1
where sampling started at 10:55 a.m. April 7 (see Table 2).

8

SECTION 8

 Table 2: Samples from Borehole 1 on the #22 Headgate section, collected 4/7/10

There were four independent sources of gases exhausting out the Bandytown fan. After the
explosion, some methane, carbon dioxide and oxygen-depleted air entered the mine from the
large sealed area south of Old North Mains. Damage to ventilation controls caused the air
current ventilating the seals to reverse from seal set 8 through 15, and this ventilation now went
to the Bandytown fan. Each seal set had water traps, with pipes in at least one seal per set,
installed for water control through the seals. A total of 18 pipes existed in seal sets 8 thru 15,
each pipe having diameters of approximately seven inches. The water traps in most of these
pipes were found empty during re-entry, and they were likely empty immediately after the
explosion due to the pressure wave from the explosion. The gradual decrease in the barometer
(Figure 4) would cause expansion of the gases in the sealed area, and the water pipes provided
sufficient path for the gas to escape in the ventilation to Bandytown. In addition, the pressure
drop to the Bandytown fan would cause airflow into some of the pipes in seal sets farther away

9

SECTION 8

 from the shaft, such as seal sets 8, 9 and 10, and flow out the remaining seal sets, especially 14
and 15. These air currents would affect the northeast segment of the sealed area. It is estimated
that about 6,400 cubic feet per minute vented from this large sealed area to the Bandytown fan
during the several hours after the explosion. Analyses of air samples from this sealed area
indicate that it contained methane, carbon dioxide and oxygen-depleted air (see Table 3:
Samples in the sealed area).

Table 3. Gas samples from sealed area taken 9/14/11

Another gas source includes the products of combustion, which include methane, hydrogen,
carbon monoxide, carbon dioxide, acetylene, ethylene and ethane. The mine normally liberated
methane and carbon dioxide from the mined coal and the gob areas. A fourth source of gas was

10

SECTION 8

 the increased liberation of methane from the floor cracks, and during the investigation it was
found that the gas from the floor cracks contained some ethane (see Table 1). All these gases
mixed and left the mine through the Bandytown fan.

Efforts to use the ethane to methane ratio and determine how much of the methane coming out
of the mine was from the floor cracks on the longwall face proved futile. Three gases that are
combustion products, methane, carbon dioxide and ethane are also produced from other sources.
Ethane is also contained in the floor cracks with methane. Carbon dioxide is also found in the
mine returns and in increased amounts behind seals. Methane was normally liberated in the
mine, and, in this instance, was entering the mine through the floor cracks. The primary gases in
the mix that were only products of combustion were hydrogen, carbon monoxide and ethylene.

Figure 4: Showing barometer trend and methane at the Bandytown fan

Sampling of these gases did not start until 5 ½ hours after the explosion and some early data was
not available; but, even had the data been available, the variables could not be isolated with
confidence to make calculations that would allow the investigators to separate the sources of the

11

SECTION 8

 gases. Some had multiple sources, and only one measurement was available after they all
mixed.
The trends of the methane, ethane and carbon dioxide varied significantly at 6:40 a.m. April 6.
An air sample from Bandytown contained 1.08% methane as well as increased ethane and
carbon dioxide. The increased carbon dioxide could only come from combustion or the sealed
area. No one was in the mine at that time. The barometer had declined some additional
amounts (see Figure 4 for the pressure trend line) and could have been a factor. The increased
methane and ethane could come from either combustion or additional liberation from the floor
cracks. Other minor fluctuations occurred in the gas trends at intervals on April 6 while no one
was in the mine.

Figure 5: Showing methane trends with time

The last air samples taken by a federal inspector at Bandytown return, on January 20, 2010,
contained an average of 0.182% methane and 0.11% carbon dioxide in 374,893 cfm of air. That
same percentage of methane, 0.18, was the average of samples collected November 3, 2009; the
air measurement then was 448,200 cfm. The quantity of air exiting the Bandytown fan is
probably less than that measured by inspectors and examiners, but the percentage of methane

12

SECTION 8

 has been consistent at 0.18%, even though both measure different air quantities. We assume
that the normal amount of methane from Bandytown fan was about 0.18 percent. The
investigators do not believe that more than 315,000 cfm was exhausted after the explosion.

Figure 6: Barometer trend on January 4, 1997

It is believed that an increase in methane liberation from floor cracks behind the shields
accumulated along the space behind the shields and was not moved away and diluted by the
ventilation system. The excess methane observed coming from the mine after the explosion was
from the combustion products, the floor cracks, and some from the large sealed area. The air
analysis from samples collected at the shaft indicated that the methane content in the air reached
near normal levels by 9:30 a.m., April 6, 2010, which was 18½ hours after the explosion. As the
recovery work progressed the gases remaining began to be flushed out, increasing the methane
content at the shaft until all areas were re-ventilated (see Figure 5).

13

SECTION 8

 8.2

Coal Dust

Coal Dust as Fuel for the Explosion

The methane ignition resulted in sufficient force to disperse and ignite coal dust. The methane
source for this explosion was much larger than the minimum amount of methane determined in
experimental work. Research has found that a small gas ignition of 84.5 cubic feet of methane
diluted to 6.5% methane in air is sufficient to lift coal dust and propagate a dust explosion. An
explosion pressure of 2.5 psig is enough to initiate a coal dust explosion where sufficient dust,
such as float dust, is present.8

The ensuing coal dust explosion generated sufficient combustible gases, such as hydrogen,
carbon monoxide, methane and ethylene, to overwhelm the gas detectors at various locations in
the affected area.

A mine explosion is not a single event where forces radiate out uniformly from the explosion
source. The static pressure of an explosion varies with the speed of the expanding heated gases
of combustion and the reflections caused by the geometry and any obstructions in the entry.
Obstructions increase turbulence and improve mixing of the gases or dusts, thereby affecting the
explosion dynamics. The explosion speed may vary depending on the available fuel and
concentration distribution as it spreads from the ignition source. The forces reverse direction at
least once, and sometimes more often, leaving damage with conflicting information. It is
difficult to trace the exact path of the explosion and careful detailed examination of all the
evidence is essential.

The hot gas expansion due to unconfined burning of non-uniform gas-air mixtures produces
eddies that will grow while the flame front searches out and follows and ignites along the
flammable boundary. Some gas could have been in the tailgate entries or against the gob along
8

Nagy J. The Explosive Hazard in Mining, Bureau of Mines IR 1119

14

SECTION 8

 the tailgate entries, layered due to the low velocity there. This could have added to the roof
layer flame propagation, either as a continued gas explosion or hybrid gas in combination with
coal dust in the tailgate, increasing in velocity as it moved outby consuming available fuel that
may have been dispersed from rib and roof surfaces by local wind forces generated by the hot
gas expansion process.

Rock dust had been applied in the mine, and the available incombustible content was
instrumental in slowing the explosion propagation (see Section 6.4 for a discussion of rock
dusting practices). The amount of incombustible content to stop a dust explosion depends upon
the size of the dust particles. Sixty-five percent incombustible is necessary to inert average mine
coal dust, but because of the fine dust particle sizes created by modern mechanical mining
processes, 80% incombustible is required.9,10

Monitoring the amount of rock dust applied to ensure compliance is an on-going problem in
mining operations. From an enforcement perspective there are not many tools for use in
evaluating adequacy of rock dust applications. Inspectors must first determine whether rock
dust has been applied, which can easily be determined by observation. Then the inspector must
determine if the amount of rock dust present meets the requirements, which is most difficult.
Even the most experienced inspectors cannot regularly discern whether an area has 50% or 70%
incombustible, and, for this reason, federal inspectors collect dust samples. Generally, citations
for visible determinations of non-compliance are supported by dust samples collected at the site
violated. There is a period of several days between collecting the sample and when the results
are available. As a result, most judgments, or determinations of the rock dust adequacy for
enforcement of the regulations or evaluations by mine officials or employees, depend upon
visible evaluation. WVOMHS&T inspectors relied on visible observations to determine
adequacy.

9

Cashdollar, K. L. et al Post-Explosion Observations of Experimental Mine and Laboratory Coal Dust
Explosions, NIOSH Pittsburgh, Pa.
10
Cashdollar, K. L. et al Recommendations for a New Rock Dusting Standard to Prevent Coal Dust
Explosions in Intake Airways, Report of Investigations 9679 NIOSH Pittsburgh, Pa.

15

SECTION 8

 Dust accumulations from operations

Airborne dust is produced at many operations in the mine. Continuous mining machines and
longwall shearers generate the most. Shuttle cars raise dust in suspension as they travel the
haulage routes, such as large trucks do on dusty roads. This airborne dust is carried by the air
currents and settles out towards returns. Transfer points on conveyor belts where coal moves
between belts are serious dust generating locations and need constant attention to remove
accumulations and re-apply rock dust. This dust enters the air stream and is deposited
downwind.

In summary, rock dust needs to be applied in all the dust-exposed areas on a regular basis to
maintain incombustible content at the proper level. Areas need rock dust applications on a
repeated basis, and, in some cases, trickle dusters are used to continuously apply rock dust in the
immediate returns from longwall shearers and continuous miners. There is no evidence this was
done at UBB. The success of the rock dusting program depends on organization, planning and
inspections. The records at the mine indicate that inspections regularly identified locations that
needed rock dust applied. Interviewed witnesses indicated a perceived deficiency in rock dust
applications at various locations. Shortly before the explosion, pre-shift log entries indicated the
need for additional rock dusting along active conveyor belts (see Table 4).

Based on the available information and records, it is believed that some areas would be rock
dust deficient at any given time. These deficient areas and areas with less than 80%
incombustible provided sufficient fuel for the dust explosion. Mines should be proactive in the
rock dust program instead of waiting until rock-dust-deficient locations have been identified.

The explosion started in the tailgate entries where return air from the longwall shearer provided
the opportunity for airborne coal dust to travel and settle leaving deposits of float coal dust in the
tailgate entries. This dust would settle on all mine surfaces, including elevated surfaces such as
ledges on the coal ribs. More rock dust is necessary to prevent explosion propagation if heavy
float dust deposits exist with a substantial amount of these deposits on elevated surfaces near the

16

SECTION 8

 roof. Coal dust on elevated surfaces, such as coal ribs, is more easily suspended by air
turbulence.11,12

Table 4: Pre-shift log entries for the month preceding the explosion indicate the conveyor belts needed
rock dusting /cleaning, but the condition was not confirmed to have been fully corrected.

Direct observations about the condition of the tailgate entries were available in statements made
by company mine examiners. Mine Examiner Charles Semenske stated the tailgate area was not
well rock dusted, but it did not look that bad. It didn’t look as bad as it did outby.13 The
WVOMHS&T inspector who was conducting an inspection at the mine had not traveled the
tailgate entries in some time so he could not provide direct observation of the conditions. The
#3 through #7 entries outby the face were intakes requiring 65% incombustible content, under
the current regulation; after the face was passed they became returns where 80% incombustible
was required. Airborne dust generated at the shearer was deposited in these returns. The
11

Sapko, M.J., Weiss, E.S., Watson, R. Explosibility of Float Coal Dust Distributed over a Coal-Dust
Substratum, NIOSH, Pittsburgh, Pa.
12
Nagy, J., Mitchell, D.W., Kawenski, E.M, Float Coal Hazard in Mines, A Progress Report, Bureau of
Mines RI 6581
13
Interview of Charles Semenske, October 26, 2010, page 17 & 57

17

SECTION 8

 shearer generates a great deal of dust, and an essential part of respirable dust control is
ventilation to dilute and carry the dust away.

On March 9, 2010, ventilation was found by a federal inspector going outby in #5 to #7 entries,
a violation to the plan. During this time, float dust would have been deposited outby in these
entries. Rock dust sample analyses indicating incombustible content in these entries outby the
longwall were not available. Available records indicated that the tailgate area outby the
longwall had not been sampled during the routine sampling procedures by MSHA, and
testimony established that it had not been re-rock dusted since the area had been developed
many months before the explosion.14

Rib Dust Accumulations

WVOMHS&T investigators had noticed the accumulation of fine coal dust at various locations
on the coal ribs as the investigation continued, and decided to investigate the character of the
dust. In August 2011, about 1½ years into the investigation, WVOMHS&T investigators
collected dust samples from the ribs in the area of the #21 Tailgate entries outby the longwall.
The coal seam was divided into two benches by a shale parting. This parting and the difference
in strength of the two coal benches influenced the manner in which the coal seam spalled.15
Since the top bench of coal was more friable, a profile of ledges was created at and below the
parting on the bottom bench where the fine coal from the top bench collected. The amount of
fine coal particles, less than 200 mesh, was influenced by the amount of fusain parting material
in the top bench. Fusain is extremely soft and crumbles readily into fine, soot-like powder and it
closely resembles charcoal in terms of both chemical and physical properties. Laboratory tests
revealed that these deposits were low in incombustible content and the dust was explosive,16 and
the investigators concluded that rib spalling had occurred prior to the explosion. These entries
14

Interview of Charles Semenske, October 26, 2010, page 17 and Interview of Harold Lilly, August 24,
2010, page 67.
15
The weight of the strata above the coal can cause fracturing, spalling, and cavitation of coal ribs if the
strength of the coal is less strength than the roof or floor strata
16
Report from NIOSH see Appendix 8.2-2

18

SECTION 8

 were developed 1½ years before the explosion, and any fine coal accumulations up on the rib
would have been available for dispersion by the forces of the methane ignition (see Appendix
8.2-1). Visible deposits of coked coal were found on the roof and ribs in the tailgate entries near
their junction with the longwall face, indicating that coal dust burned in this area (see Figure 7
depicting the locations of visible macro-coking).

Figure 7: Locations where large amounts of coke and burning were found during the investigation.

The coal dust explosion could be traced from the tailgate area. It propagated at various speeds
determined by the amount of fuel and incombustible rock dust, and moved east and north from
the tailgate entries. The flame speed of a coal dust explosion decreases with its incombustible
content (for uniform mixtures of coal and rock dust). For example, Cybulski documents a flame
speed of 385 feet per second at 55% incombustible and about 144 feet per second at 65%
incombustible with 85% minus 200-mesh coal dust.17

17

Cybulski, W. G. [1975] Coal Dust Explosions and their Suppression, Translated from Polish, Warsaw
Poland, National Center for Scientific Technical and Economic Information, NTIS No. TT-54001

19

SECTION 8

 Regardless of the precise location of the ignition source the WVOMHS&T investigators believe
they have sufficient evidence to determine that the explosion propagated from the tailgate area
as a coal dust explosion (see Appendix 7.9-5). Both gas and dust explosions will leave similar
results, and it sometimes is quite difficult to determine which occurred. A gas explosion will
suspend coal dust in the air and pyrolize some of the dust in an oxygen deficient atmosphere,
leaving coke particles and coke residue even though the dust did not add significantly to the
explosion strength. A hybrid explosion,18 where coal dust and methane both contribute to the
explosion, will leave coke particles and coke residue. Methane in a hybrid mixture will more
quickly consume most of the oxygen before the coal particles heat and release their volatile
matter to the combustion process depleted of oxygen and contributing little to driving the
leading edge of the flame propagation. Therefore, the degree and rate of the combustion of coal
dust in a hybrid mixture depends on the concentration of methane and availability of oxygen. A
propagating coal dust explosion leaves visible coke deposits in far greater quantities than gas or
hybrid explosions.19 The amount of visible deposits of coke found supports a determination that
coal dust burned and propagated the explosion, (see Figure 7).

Fires and visible coke in place at various locations in the explosion area evidence the extent of
flames (see Section 7.9, Map 13). The presence of visible coke is influenced by the flame
speed. In weak and slow explosions the most vigorous combustion occurs near the center of the
mine passageway and little or no combustion takes place in the boundary layer along the roof
and ribs.20,21 Methane in explosive quantities could not have existed in all these areas prior to
the explosion because many were ventilated by strong intake air currents.

Further evidence of a coal dust explosion is found by analyzing the air samples collected after
the explosion in Borehole #1. The first of these samples was collected at 10 a.m. on April 7,
2010, while the ventilation was stagnant in the area and the gas mixture was undisturbed. The
18

Nagy, J. The Explosive Hazard in Mining, Bureau of Mines IR 1119, page 33
Rice, G. S. et al Coal Dust Explosion Tests in the Experimental Mine 1913 to 1918 Inclusive, Bureau of
Mines Bulletin 167
20
Hartman, I., Studies on the Development and Control of Coal Dust Explosions in Mines, Bureau of
Mines IC 7785, page 12
21
Richmond, J.K., Liebman, L., Miller, L.F., Effects of Rock Dust on Explosibility of Coal Dust, Bureau of
Mines RI 8077, page 4
19

20

SECTION 8

 Jones-Trickett Ratio22 was developed and tested on many experimental mine explosions to help
differentiate between gas or dust explosions. When the ratio is between 0.4 and 0.5, methane
was the primary fuel; values between 0.5 and 0.9 indicate that the fuel is primarily coal. We
used the component of the gases collected and analyzed by MSHA from Borehole #1 to
calculate the Jones-Trickett Ratio, and the results indicate the combustion products were from a
coal dust explosion (see Table 2 showing the results of this sampling).

Microscopic evidence of coal as fuel

Other evidence of a coal dust explosion exists. Coal dust particles that participate in dust
explosions are subject to rapid heating causing them to pyrolize and to soften or melt into a
plastic phase. The surfaces of these burned residues are rounded and smooth in contrast to the
angular features and sharp edges of unburned coal or coke. The burned particles are
pockmarked with blowholes, and some are blown into large cenospheres (empty hollow
spheres).

Approximately 65 samples of impacted dust taken at various locations by WVOMHS&T (see
Figure 8 for locations where samples were taken) were analyzed by NIOSH using a Scanning
Electron Microscope. Nearly all 65 samples23 demonstrate the presence of particles that have
been heated, leading to the conclusion that coal dust participated in an explosion (see Figure 9
for representative images of these particles).

22

Morris, R The Jones-Trickett Ratio-a Worthwhile management Tool, Journal of The South African
Institute of Mining and Metallurgy May 1997 Page 149
23
A few samples from the shield cylinder legs on the Longwall lacked distinct cenospheres

21

SECTION 8

 Figure 8: Locations where impacted dust was analyzed for cenospheres. All locations had relatively
abundant cenospheres, except for the Longwall, where cenospheres were present but less abundant.

Flame traveled throughout much of the area of the longwall entries and inby. MSHA dust
samples collected during the investigation to determine the incombustible content were routinely
analyzed for amounts of coke using the alcohol coke test. Research has found that coke found
by such tests indicate the passage of flame in the mine entry,24 and the samples indicate that
flame traveled throughout most of the area, including the tailgate entries to about 1,500 feet inby
the longwall face. The flames traveled into #22 Headgate section, the #8 North and #9 North
Areas, and the Glory Hole Area. Fires were started in the North Glory Mains near survey
station 20057. Evidence of flame exists in the tailgate entries near the junction of North Glory
Mains. Heated materials were found on the longwall face near shield 62 that probably came
from burning methane behind the shields before, during or slightly after the main explosion in
the tailgate. Flame evidence exists in the North Mains to about where the entries make a turn to
the left. Heated gases deposited soot stringers beyond this point.

24

Man, C. K. et al Determining Flame Travel Measurements From Experimental Coal Dust Explosions,
NIOSH, Pittsburgh, Pa.

22

SECTION 8

 Additional information about cenospheres in UBB samples can be found in Appendix 7.9-1,
Appendix 8.2-3, and Appendix 8.2-4.

Example A

Example B.

Example C.

Figure 9: These photographs taken under a scanning electron microscope depict the type of burned
residues present in the samples.

23

SECTION 8

 9

Conclusions and Recommendations

9.1

Conclusions

9.2

Recommendations

9.3

Acknowledgements

 9

Conclusions and Recommendations

9.1

Conclusions

1)

The explosion at the Upper Big Branch Mine on Monday, April 5, 2010, occurred at

approximately 3:02 p.m. The explosion was the result of methane accumulating in the #21
Longwall Tailgate T-split area and in the gob behind the longwall shields, which was then
ignited.
2)

There are two possibilities for the ignition. It is possible that sandstone roof rock

falling behind the longwall shields ignited gas by rock-on-rock impact or rock-on-steel impact.
However, the more likely cause is that it occurred on the longwall as the shearer was cutting
sandstone roof, igniting gas by steel-on-rock impact along the shearer bits producing hot
smears as it cut sandstone roof.
3)

Approximately 90 seconds elapsed from the time the shearer was de-energized until

the time of the explosion.
4)

The actual explosion most likely began in the gob behind the longwall shields,

between shield 173 and shield 117. Explosion forces propagated along a flammable gas fringe
behind the shields north to approximately shield 65 to 55, where access to fuel and space were
restricted and the forces were diverted into the longwall where they subsided to extinction.
Explosion forces also propagated west/southwest through the gob, exiting the gob 300 to 400

1

SECTION 9

 feet west of the longwall shearer. Additionally, explosion forces from the gob propagated
south/southeast entering the longwall a short distance north of the shearer and traveled across
the shearer as it exited the longwall and entered the #7 entry of the #21 Tailgate. From these
two locations in the #21 Tailgate, the explosion transitioned into a coal dust explosion which
continued propagating east and west and throughout an extensive area of the mine.
5)

Coal float dust, combined with fine coal dust derived from rib spalling, is believed to

be the principle source of fuel for the propagating explosion.
6)

An airflow restriction, approximately 45 feet inby the shearer in the #7 entry of #21

Tailgate, caused by roof caving associated with the advancing longwall would have allowed
any gas migration into that location to build up. A small amount of air could move over and
around the fall, but the majority of the air did not.
7)

Methane, mixed with other gases, is emanating from a source beneath the mine floor.

A combination of overburden cover depth, mine floor rock strength, and local fracturing
(natural and induced) influence where connecting pathways to gas are intercepted during
mining. Prior gas events and the April 5, 2010, explosion appears correlated with local
thickening of an underlying coal seam located 10 to 15 feet below.
8)

The removal of hazards and violations identified during required mine examinations

were not corrected in a timely manner.
9)

There was no indication that the #21 Tailgate entries west of the #21 Cross-over had

been rock dusted since the Longwall began production in September 2009. Based on
testimony and available information there is no evidence that additional rock dust was applied
after initial development throughout this extensive area.

2

SECTION 9

 9.2

Recommendations

Additional Explosion Defenses
1)

The lines of defense to eliminate mine explosions are adequate ventilation, detection of

explosive gases, removal of ignition sources, cleanup of fine coal dust accumulations and rock
dusting. In other words:
a)

prevent methane accumulations or find the accumulation before it becomes a problem.

b)

prevent the ignition if an accumulation occurs undetected.

c)

prevent a coal dust explosion if an undetected accumulation of methane becomes ignited.

All these defense mechanisms failed at UBB.

Additional defenses are needed to prevent propagation of a methane explosion into a coal dust
explosion. Explosion barriers have been studied for years and have been found to be effective
in stopping an explosion.

Investigators found evidence that a water barrier just west of #22 Cross-over stopped the
explosion from propagating further inby in the #21 Headgate entries. The first arrival of
explosion forces entered through a water-filled dip and then passed over a gob pile which
reduced the opening height to about 4 feet. The result was recorded by impacted mud deposits,
instead of impacted dust, and by the rapid decay in heat and pressure forces inby. The
explosion subsided to extinction between breaks 31 and 35 of #21 Headgate, and further
evidence of an explosion was not found as far west as could be safely traveled and examined.

Research is needed to demonstrate the practical application of water barriers, rock rubble
barriers and other explosion mitigation strategies as supplemental protection, with generalized
rock dusting, to prevent explosion propagations. It is recommended that NIOSH renew their

3

SECTION 9

 research into the subject of barriers and develop systems that can be used in operating coal
mines.

Ventilation Systems
1)

WVOMHS&T currently have insufficient statutory language to regulate the way that

coal mines are ventilated.
2)

Coal operators must take a more proactive approach to the ventilation of each coal mine

under their authority and responsibility. The industry has taken a step back over the years in
proper planning and preparation as long-term plans are developed. The operator must once
again lead the way in doing the hard and extensive work necessary to properly ventilate each
mine. Registered Professional Engineers can help bridge the gap between prudent planning and
maintenance of mine ventilation. Mine managers and regulators must allow the Professional
Engineer to do his/her job and should not be so quick to disregard the engineer’s professional
judgment, which he/she is charged by charter to do in order to protect public safety.

3) A closer look must be taken concerning the use of belt air during longwall mining to
assure that the most effective means are being utilized to maximize ventilation to the
longwall face and the active gob areas including the tailgate T-Split (i.e., where the
longwall face intersects its principle return air course).

4)

The T-split is a critical location where longwall ventilation can become blocked,

which is what happened at UBB. Additional measures to support the roof at this location
are prudent and recommended. Many operators are already doing this. In addition, should
the T-split become blocked, there should be supplemental precautionary measures that are
initiated to ensure the ventilation system is keeping gas mixtures properly diluted.

5)

The excessive use of airlock doors along the ventilating current must be minimized to

assure that ventilation is not compromised. Although this is currently addressed in the
West Virginia Code, it requires clarification with stricter guidelines for enforcement.

4

SECTION 9

 6)

Critical air splits in the mine should be adequately monitored at strategic locations to

give warning of unusual or dangerous amounts of methane, carbon monoxide or other
harmful gases. This system should alert the operator if an air reversal or a drastic change
occurs in the ventilating air volumes. These devices should give warning to a central
location on the surface so workers in the affected areas can be alerted to hazardous
conditions.

7)

Mine management should review the mine ventilation, roof conditions and any other

specific hazard to the mine, such as inundation potential, and organize and schedule mine
examinations to ensure conformance to the law and regulations, being specifically attuned
to the potential hazards at the mine. Examiners should be trained in the examination
schedule and locations.

Mine Examinations
1)

Additional training must be conducted to assure that the operators and those entrusted

by the operator to make the required safety exams in the mines be adequately trained in the
performance of their duties. Training to clearly identify hazards and the corrective actions
to eliminate those hazards must become standard throughout the mining community and
immediately replace the routine haphazard practices in place today. This training should
be an extensive part of the foreman continuing education classes required by the
WVOMHS&T.

2)

The system for mine examiners to record their findings for review must be overhauled

to allow for a more thorough recordkeeping and review process of noted conditions. This
would enhance the ability of coworkers who are charged with correcting these hazards and
violations to properly evaluate their findings and develop a precise plan to correct the
hazards and violations. The current fireboss books being used need updated or redesigned

5

SECTION 9

 so that the hazards recorded can be easily tracked to assure they have been corrected
promptly.

3)

Every person who is given the task of making required mine examinations must use a

multi-gas detector that will log or record their findings during the course of their
firebossing assignments. A means must be available for their detection device to log its
readings and be made available to be printed out or recorded for review and scrutiny by
the operator and the regulatory agencies. This record must be kept like any other required
record for a designated period of time and made available for review upon request by the
proper authorities.

4)

All continuous miners and longwalls should have methane monitors maintained in a

workable condition at all times. If the methane monitor becomes inoperative, the machine
should be taken out of service. The methane monitors should be calibrated every 15 days
in a known gas mixture and a record kept of these calibrations.

5)

Individuals involved in the day to day decision making at the mine must be held

accountable regardless of their title. The mine foreman is the highest ranking official that
current state law addresses.

Rock Dusting
1)

Each mine should have a means to track the amount of rock dust that is delivered to

the mine and when and where it is applied underground. This recordkeeping should be
done in a manner that will allow the operator to easily review and track not just the amount
of rock dust used but also monitor the frequency in which the rock dust is being applied to
the mine surfaces. This record should be made available upon request.

2)

Rock dusting should be maintained to 80% total incombustible content within 500

feet of an active section or any ignition source within 500 feet downstream of mine seals.

6

SECTION 9

 3)

Each mine should conduct its own rock dust surveys to assure compliance with the

laws pertaining to total incombustibility of 80%. This should also include periodic
analysis of the rock dust being used to assure that the quality of the dust being used is
compliant. The way we currently sample and apply rock dust needs to be modernized,
using the latest research knowledge and most appropriate sampling and analyzing methods
available. A record should be kept of testing results and made available upon request.

4)

Clarification of some basic rock dust and rock dusting facts should be communicated

to the mining community regarding adequate rock dusting methods and the need for
increased frequency of rock dusting. A better understanding of how a dust explosion
begins and then propagates would help the average miner to understand why it is crucial to
conduct adequate exams and to rock dust properly and would encourage workers to be
more diligent concerning rock dusting the work place.

Emergency Responders

1)

Hazards associated with new urethane glues, sealants, foams and fine silica dusts,

including cenosphere fragments which are heated at high temperature and dispersed during
a mine explosion, are not well understood. In the event another emergency response and
investigation of an underground coal mine as extensive as UBB were to occur, it is
recommended that steps be taken to examine the health risks of working in such an
environment. Organic respirators or other breathing protections should be made available
for mine rescue personnel and investigators who work underground.

7

SECTION 9

 9.3

Acknowledgements

Special recognition is given to Timothy Blake of the 22 Tailgate crew. The WVOMHS&T
employees, the coal miners and citizens of the State of West Virginia would like to thank you
for your heroic efforts on April 5, 2010. Your efforts to assist your coworkers will always be
remembered.
The West Virginia Office of Miners’ Health, Safety and Training wishes to acknowledge and
thank all of the family members and mine employees past and present who participated and
assisted this agency and followed the proceedings of our investigation. Special recognition is
due to those who appeared at the hearings and for the forthright manner in which they gave
their testimony and their forbearance while spending many hours traveling or waiting to appear.
We would like to thank the citizens of the Whitesville/Big Coal River area for all the continued
support, prayers and hospitality that they extended to us throughout our stay in your
community. We appreciate the support that the Whitesville Fire Department and EMS showed,
as well as all of the other area emergency medical services and local fire departments.
We especially want to thank all the mine rescue teams for their support and assistance during
the rescue and recovery events.
Our thoughts and prayers continue to be with James K. Woods survivor of the 22 Tailgate
crew.
The following groups and individuals were involved with the events or the investigation
surrounding the disaster that occurred April 5, 2010:
The MSHA Upper Big Branch Investigative team; Norman Page, lead investigator
Performance Coal Company
United Mine Workers of America (UMWA) led by Max Kennedy
Governors Independent Investigative Panel led by Davitt McAteer
National Institute of Occupational Safety and Health Administration (NIOSH)
Governor Earl Ray Tomblin
Senator Joe Manchin III (Former Governor)
The Governor’s Office Staff

8

SECTION 9

 West Virginia Homeland Security
Southern West Virginia Community College Task Force One
West Virginia State Medical Examiners’ Office
Local and County government agencies
West Virginia State Police
Raleigh and Boone County Sheriff Departments
Raleigh and Boone County Commissions
West Virginia Department of Health and Human Resources
West Virginia Regional Response Team
C.A. Phillips, Director, WVOMHS&T
Eugene White, Deputy Director, WVOMHS&T
WVOMHS&T Upper Big Branch Investigative Team, William Tucker lead investigator
Mr. Barry Koerber, Assistant Attorney General, State of West Virginia
Ronald Wooten, (former Director, WVOMHS&T)
Terry Farley, (former Administrator, WVOMHS&T)
Monte Hieb, Chief Engineer, WVOMHS&T, your tireless efforts in leaving no stone unturned
is appreciated. Your open minded approach since the onset of this tragedy has kept the focus
on finding the facts surrounding the events which occurred April 5, 2010.
Jack E. Tisdale, consultant. Your expertise in ventilation, explosions and explosion prevention
were integral to the development of this report.
Jeff Kohler, Michael J. Sapko, Jürgen Brune and Isaac A. Zlochower, NIOSH. Thank you for
rendering technical assistance to this investigation.
University of Kentucky; Mr. Cortland F. Eble, Mr. Henry Francis

9

SECTION 9

 10

List of Appendices

APPENDIX 1
Definitions
Errata

APPENDIX 2
Geology and Mining Conditions
2.2-1 Geology
2.2-2 Petrographic and X-ray diffraction analysis of sandstone roof from the Longwall

Communication and Tracking System
2.5-1 The Approved Communication and Tracking Plan
2.5-2 Violation and extensions
2.5-3 Time drift correction: Pyott-Boone PC clock
2.5-4 Calculations for 108 second delay: Pyott-Boone

1

SECTION 10

 APPENDIX 3
Incidents of Prior Gas Inundations
3.2-1 Report of a Methane Ignition at UBB, January 4, 1997

APPENDIX 4
Origin and Path of the Explosion
4.1-1 Longwall production reports
4.3-1 Effectiveness of a water barrier in stopping the explosion in the inby reaches of
#21 Headgate

APPENDIX 5
Chronology of Events
5.1-1 List of Victims

APPENDIX 6
Regulatory Compliance
6.2-1 Belt Books

APPENDIX 7
Structure of the Joint Investigation
7.2-1 Individuals who exercised 5th Amendment Rights

2

SECTION 10

 Flames and Forces Mapping
7.9-1 Sampling and Analyzing of Mine Dusts
o Exhibit 1- WVOMHS&T impacted dust samples: Summary of test results
7.9-2 Using Roof Pans as Indicators of Wind Pressure and Heat
o Exhibit 1- Format for describing condition of roof pans
7.9-3 Identifying Final Forces
7.9-4 Using Breached and Deflected Structures to Interpret Explosion Forces
7.9-5 Forces and Heat on the Longwall
o Exhibit 1- MSHA shield inspection data
o Exhibit 2- Mallett Technology Report, executive summary; CFD analysis of
shield signs and CIU E-stop buttons.
7.9-6 Explosion Pressures
7.9-7 West Virginia Flames and Forces Map---Eight (8) sheets (see Appendix 9)

APPENDIX 8
Mine Gases
8.1-1 Tailgate Intersection

Coal Dust
8.2-1 Mine Dust and Rock Dust
o Exhibit 1 Petrographic Analysis of Coal Samples
8.2-2 Particle size analysis and explosibility analysis on rib dust samples
8.2-3 The Imaging of Post-Explosion UBB Samples using a Scanning Electron
Microscope
8.2-4 The Imaging of Additional Post-Explosion UBB Samples using a Scanning
Electron Microscope

3

SECTION 10

 APPENDIX 9
West Virginia Flames and Forces Maps
Sheets 1 thru 8: D-size (34” x 22”)

4

SECTION 10