CALIFORNIA AIR RESOURCES BOARD 
 
 
 

 

Determination 
of Total 
Methane 
Emissions from 
the Aliso Canyon 
Natural Gas 
Leak Incident 

October 21, 2016 
 
 

This  report  documents  the  California  Air  Resources  Board 
(ARB)  staff’s  determination  of  the  total  methane  emissions 
from  the  Aliso  Canyon  natural  gas  leak  incident  and  the 
amount  needed  for  full  mitigation  of  the  climate  impacts. 
The mitigation is expected to be accomplished with projects 
funded by the Southern California Gas Company. The report 
summarizes  the  various  efforts  by  ARB  and  others  to 
measure  methane  emissions  from  the  Aliso  Canyon  natural 
gas  leak  incident,  and  how  they  were  used  to  estimate  the 
total methane emitted from the incident. The total amount 
of methane  needed to fully mitigate the climate impacts of 
the leak is 109,000 metric tons. 

 

 
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Table of Contents
Summary ..................................................................................................................................................... 1 
Background ................................................................................................................................................. 2 
Quantifying the Amount of Methane Emitted ......................................................................................... 5 
Preliminary ARB Estimate ..................................................................................................................... 5 
Emission Estimates from Southern California Gas Company ......................................................... 6 
Inventory Verification ......................................................................................................................... 6 
Tracer Flux Ratio Study ..................................................................................................................... 7 
Estimates by Research Organizations .............................................................................................. 12 
Refinement of the Preliminary ARB Estimate ...................................................................................... 15 
Final emission estimates and amount of methane that needs to be mitigated ............................... 27 
 
 

 
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Summary
This report documents the California Air Resources Board (ARB) staff’s determination of
the total methane emissions from the Aliso Canyon natural gas leak incident and the
amount needed for full mitigation  of the climate impacts. The leak at the Southern
California Gas Company’s (SoCalGas) Aliso Canyon natural gas storage facility in Los
Angeles County was reported on October 23, 2015 and controlled on February 11,
2016, and has been described as the largest documented leak of methane in the United
States1. In response to the leak, Governor Brown issued a proclamation2 that directed
ARB to prepare a program, to be funded by SoCalGas, which will “fully mitigate the
leak's emissions of methane”. SoCalGas has indicated that it is committed to mitigating
the climate effect of the leak, and is ready to work with ARB staff to develop the
appropriate programs to do so3.
The Aliso Canyon methane release was caused by an uncontrolled breach in the
natural gas storage infrastructure and occurred outside the envelope of instruments put
in place to measure the flow of natural gas at the facility. The size of the leak was
therefore initially unknown, and could only be evaluated qualitatively using available
ambient air measurements in the vicinity of the leak. Subsequent to the leak being
controlled, the scientific research community and SoCalGas have provided different
methane emission estimates based on a variety of measurements including ambient
measurements, remote sensing, and an inventory accounting method. ARB also
provided a preliminary estimate based on downwind flights that measured flux rates at
various times during the leak. The Los Angeles basin has a uniquely dense network of
methane measurement efforts and instrumentation that allow for several different
quantification methods to be used to estimate total amount of methane released. The
additional information from this network was used to inform and corroborates ARB’s
final estimate. This report provides an overview of the various efforts, and describes
ARB’s updated approach to estimate the total methane emissions from the Aliso
Canyon natural gas leak incident.
ARB’s updated estimate indicates that the incident resulted in a total emission of
99,650 ± 9,300 metric tons of methane. To fully mitigate the leak, as directed by the
                                                            
1

Conley, S., et al. (2016). Methane release rates from a single leak were nearly double that of the entire
rest of the Los Angeles region. Science, 351 (6279), pp. 1317-1320, DOI: 10.1126/science.aaf2348
2

Governor’s Proclamation of a State of Emergency (January 6, 2016),
https://www.gov.ca.gov/news.php?id=19263
3

Letter from Mr. Dennis V. Arriola, President and Chief Executive Officer of SoCalGas to Governor
Brown, (December 18, 2015), https://www.arb.ca.gov/research/aliso_canyon/sgc_letter_mitigation_12-182015.pdf

Page 1
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Governor’s Proclamation, the upper bound of this estimate should be used. Hence, the
required amount of methane that needs to be mitigated is 109,000 metric tons.

Background
On October 23, 2015, Southern California Gas (SoCalGas) informed the State of a
natural gas leak at its Aliso Canyon natural gas storage facility. Natural gas is
composed primarily of methane, which is a potent greenhouse gas (GHG) and classified
in the category of gases called short-lived climate pollutants (SLCPs). These types of
gases remain in the atmosphere for much shorter periods than longer-lived climate
pollutants, such as carbon dioxide (CO2); but when measured in terms of how they heat
the atmosphere, pound for pound their impacts can be tens, hundreds, or even
thousands of times greater than that of carbon dioxide4. Methane is 84 times more
potent that CO2 on a 20-year time scale (based on global warming potentials from the
Fifth Assessment Report of the Intergovernmental Panel on Climate Change5).
The leak had an extraordinary impact on the local population. The emissions were a
major public nuisance, and resulted in more than 2,300 odor complaints during the leak
from the nearby communities to the South Coast Air Quality Management District and
the relocation of over 6,800 households and several schools. The complaints included
odor, dizziness, headaches, nausea, eye, nose and throat irritation, and nose bleeds.
Concerns were also expressed about exposure to other compounds found in natural
gas, such as benzene, oil residue, hydrogen sulfide, and radon in the neighboring
communities, although measurements did not suggests levels were sufficient to be of
concern.
In addition to the leak’s many effects on local residents, the methane emissions from
Aliso Canyon exacerbated statewide greenhouse gas emissions which contribute to
climate change, a problem California has recognized and is working to address. The
estimates show that during the leak, it alone was responsible for approximately
20 percent of statewide methane emissions, which is more than double the statewide
fugitive emissions from oil and gas production (Figure 1). The ability of a single source
to materially affect total statewide emissions is a serious concern.

                                                            
4

Kirschke, S., et al. (2013), Three decades of global methane sources and sinks, Nature Geoscience, 6,
pp. 813–823. http://www.nature.com/ngeo/journal/v6/n10/full/ngeo1955.html?WT.ec_id=NGEO-201310
5

IPCC (2014): Climate Change 2014: Synthesis Report. Contribution of Working Groups I, II and III to the
Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Core Writing Team, R.K.
Pachauri and L.A. Meyer (eds.)]. IPCC, Geneva, Switzerland, 151 pp. https://www.ipcc.ch/report/ar5/  

Page 2
 

 CALIFORNIA AIR RESOURCES BOARD 
 

600

Methane Emissions
(thousand MT CH4)

500
400
300
200
100
0
Statewide Total
(excl. Aliso
Canyon)

Fugitive methane Fugitive methane Aliso Canyon Leak
from oil and gas
from pipelines
production

Figure 1: Statewide total methane emissions and oil and gas methane emissions during the leak
compared to the Aliso Canyon leak.

California is committed to reduce the emission of methane and other SLCPs. A decade
ago, the State enacted the California Global Warming Solutions Act of 2006 (Assembly
Bill 32, Nunez), which requires a return to 1990 statewide greenhouse gas emission
levels by 2020. The passage of Senate Bill 32 (Pavley) on September 8, 2016, sets an
even more ambitious target by requiring the State to reduce GHGs to 40 percent below
1990 levels by 2030. The Legislature directly recognized the critical role that SLCPs
must play in the State’s climate efforts with the passage of Senate Bill 605 (Lara), which
required ARB to develop a strategy to reduce SLCP emissions. The draft SLCP
Strategy was released in April, 2016 and outlines plans to reduce emissions of methane
by 40 percent below current levels in 2030. Subsequently Senate Bill 1383 (Lara) was
passed, which requires a 40 percent reduction in methane from 2013 levels by 2030.
On December 18, 2015, with the leak still ongoing, Mr. Dennis V. Arriola, President and
Chief Executive Officer of SoCalGas, wrote a letter to Governor Brown committing
SoCalGas to “mitigate the environmental impact of the actual natural gas released from
the leak” and “working with you and your staff to develop a framework that will help us
achieve this goal.”3
On January 6, 2016, Governor Brown issued a proclamation2 that included direction to
ARB to prepare a program, to be funded by SoCalGas, that will “fully mitigate the leak's
emissions of methane”, “be limited to projects in California”, and “prioritize projects that
reduce short-lived climate pollutants”. This program, with its focus on the leak’s climate
impacts, represents one facet of a comprehensive response by State and local
Page 3
 

 CALIFORNIA AIR RESOURCES BOARD 
 

agencies to the leak and its short- and long-term effects upon the environment, and
public health and safety.
In response, ARB developed and released the mitigation program on March 31, 20166.
The mitigation program was a product of a public process that included a presentation
to the ARB Board7, as well as two periods of public comment, the second of which
followed the posting of a draft version of this program on ARB’s website. Altogether,
ARB received, and considered, more than 60 comments on the mitigation program from
the public and other stakeholders prior to finalizing the plan.
In order to quantify the methane emissions from the Aliso Canyon gas leak, State
agencies, in collaboration with scientific experts, relied on existing and new methane
measurements in the Los Angeles basin, including ambient measurements around the
well site, at nearby air monitoring towers, and using airplanes, as well as remote
sensing and satellites. These measurements allow for an estimation of the leak’s
cumulative emissions and can be used to estimate a mitigation threshold that provides
confidence the emissions will be fully mitigated.
This report summarizes the various efforts by ARB and other institutions including the
National Aeronautics and Space Administration’s Jet Propulsion Laboratory
(NASA-JPL), California Institute of Technology (Caltech), Scientific Aviation, and
SoCalGas to determine the total methane emissions from the Aliso Canyon natural gas
leak incident, provides insight on the strengths and limitations of each of the methods,
and updates ARB’s preliminary estimate. The updated estimate is then compared to the
other currently available estimates. The report concludes by providing the amount of
methane required for full mitigation of the leak.

                                                            
6

ARB Aliso Canyon Methane Leak Climate Impacts Mitigation Program, (March 31, 2016),
https://www.arb.ca.gov/research/aliso_canyon/arb_aliso_canyon_methane_leak_climate_impacts_mitigat
ion_program.pdf
7

Presentation to the ARB Board: Update on Aliso Canyon Methane Leak (February 18,
2016),https://www.arb.ca.gov/board/books/2016/021816/16-2-1pres.pdf 

Page 4
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Quantifying the Amount of Methane Emitted
The following sections compile the various methods that have been used to estimate
the amount of methane leaked, discusses the quality of each of the methods, and
updates ARB’s preliminary estimate.

Preliminary ARB Estimate
The first estimate of the Aliso Canyon natural gas leak emissions were generated by
ARB using airborne flight measurements conducted during the leak event. ARB and the
California Energy Commission coordinated 11 downwind flights by Scientific Aviation
during the leak, using a small airplane equipped with instruments that measure
methane. The plane is able to capture both the horizontal and vertical extent of the
plume by flying downwind through the plume at many different elevations. These
downwind measurements were used to make a calculation of the instantaneous
emission rate (the rate of emissions at the time the measurement was made) using
scientifically peer reviewed methods (Figure 2). However, the measurements do not
shed light on the emission rate in the periods between the flights which may have been
variable, especially prior to late November 2015 due to the activity at the well pad and
the many attempts to control the well.
An initial rough estimate of the leak was made by utilizing the airborne measurement
data, and assuming the leak rate was constant between each flight. Using this
approach, it was estimated that the Aliso Canyon incident released 5.4 billion cubic feet
(Bcf) of natural gas or 94,500 metric tons (MT) of methane into the atmosphere, which
is equivalent to emitting about 7.9 million metric tons of carbon dioxide equivalent
(MMTCO2e) (using the 20-year global warming potential for methane from the Fifth
Assessment Report5).
70,000
Methane Emission Rate
(kilograms/hour)

60,000
50,000
40,000
30,000
20,000
10,000
0
10/23

11/6

11/20

12/4

12/18

1/1

1/15

1/29

2/12

Figure 2: Updated results from the downwind flights showing hourly emission rates

Page 5
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Emission Estimates from Southern California Gas Company
On May 26, 2016, SoCalGas issued a press release that described their estimate of
leaked methane emissions from the Aliso Canyon leak incident from October 23, 2015,
until February 11, 2016, when the leaking well was controlled via the relief well8.
SoCalGas reported estimates of the total methane emissions using two separate
methodologies, which are referred to as “inventory verification” and “tracer flux ratio
study (TFR)”, that yielded estimates of 4.62 Bcf and 4.75 Bcf (84,200 MT and
86,000 MT), respectively. Below is a description of the methods used in each estimate.
Inventory Verification
The first estimate released by SoCalGas used a mass balance approach that utilized
storage field inventory estimates prior to and after the leak and an accounting of all
withdrawals and injections to infer how much methane was “missing”. SoCalGas has
stated the inventory method is not designed to calculate a leak rate, and has not
provided an uncertainty for this estimate, which is expected to be significant considering
the errors in both the inventory estimates9 and in accounting for all withdrawals and
injections. To better understand whether the result from the inventory method is
inconsistent with the other estimates, an expected error, or uncertainty, of the inventory
verification estimate is needed. If the uncertainty in the inventory estimate encompasses
the other estimates, they would not be considered inconsistent.
Inventory estimates at the Aliso Canyon natural gas storage facility were last made in
October 2014 and March 2016. Between these two estimates the reservoir has been
emptied, then filled, and then emptied again (the reservoir is typically emptied during
November – March and filled during April – October every year). The Aliso Canyon
Natural Gas Storage reservoir has a capacity for more than 167 Bcf of natural gas, of
which 86 Bcf is usable (working) gas and 81 Bcf is cushion gas. To make an inventory
estimate all withdrawals and injections are stopped for several weeks, and near steady
state pressure readings from wells throughout the reservoir are used to infer the amount
of natural gas in the reservoir. Uncertainty in calculating the inventory based on the
pressure measurements range from 3 to 10 percent10. Assuming the most conservative,
or lowest, uncertainty of 3 percent in the inventory estimates and assuming a
                                                            
8

SoCalGas (2016), SoCalGas Delivers on Commitment to Conduct Thorough Measurement of Emissions
from Aliso Canyon Leak [Press release, May 26, 2016].
https://www.socalgas.com/1443739536978/thorough-emissions-measurement-52616.pdf
9

Tutt, C. J., and Dereniewski, E. (1978), A Practical Regression Model of Pressure/Inventory Hysteresis
in Natural Gas Storage Fields, SPE 6488, doi:10.2118/6488-PA
10

Tek, M.R. (1991), Errors and Uncertainty in Inventory Verification in Underground Storage, SPE 23829

Page 6
 

 CALIFORNIA AIR RESOURCES BOARD 
 

0.5 percent uncertainty in accounting for injections and withdrawals, the overall
emissions result of 84,221 MT of methane is expected of have an uncertainty of
125 percent, or range from 0-190,000 MT. Table 1 shows the calculation of uncertainty
for the inventory method as a function of the accuracy in each of the numbers required
to make the estimate. The total uncertainty is provided as the square root of the sum of
squares of the uncertainty for each of the terms used to calculate the 4.62 Bcf.
Table 1 – A simple model to estimate uncertainty for the storage field inventory method
Inventory Method for Calculating Leak

Expected Uncertainty
Uncertainty in
each term
(Bcf)
Uncertainty

[Bcf]
Amount in Reservoir October 2014

Inventory

3.0%

Withdrawals/Injections

0.5%

167.53

5.03

Withdrawn during winter 2014-2015

-

70

0.35

Injected summer 2015

+

70

0.35

Withdrawn during winter 2015-2016

-

70

0.35

Amount in Reservoir March 2016

-

92.91

2.79

Leaked amount

4.62

Total Uncertainty (Bcf*)

5.78

Uncertainty as a percentage of estimate

125%

 
* Total uncertainty calculated as the square root of sum of squares for individual terms: 

5.03

3

0.35

2.79  

Tracer Flux Ratio Study
The second estimate released by SoCalGas utilized a tracer release method. This
technique relies on releasing a known quantity of a tracer gas under controlled
conditions near the leak site, and measuring the concentration of the gas along with
methane in downwind locations. The ratios of tracer to methane concentration and
tracer concentration to tracer source size are then used to estimate the methane
emission rate. SoCalGas contracted with Aerodyne to conduct the tracer release study
by deploying a mobile measurement platform to sample in the plume downwind from the
facility along Sesnon Boulevard. Aerodyne conducted 25 tracer measurements between
December 21, 2015, and the end of the leak on February 11, 2016. The tracer studies
consistently estimated a lower leak rate than Scientific Aviation during late December
2015 and January 2016 when both measurements were being made (Figure 3).
Assuming that the TFR technique would have resulted in the same lower emission
estimates compared to the flights prior to December 20, Aerodyne estimated a methane
emission of 86,022 ± 8,393 MT of methane.
Page 7
 

 CALIFORNIA AIR RESOURCES BOARD 
 

TFR is generally a well-practiced methodology when estimating emissions from point
sources in terrain that allows for good mixing of the tracer and the point source. There
are four important factors in the Aliso Canyon case that suggest the uncertainty in the
estimate would be larger than indicated by SoCalGas.
1. Degree of mixing of the tracer with the plume,
2. The assumption that the leak can be treated as a point source,
3. Road access downwind of the leak to capture the full downwind plume and
determine the true methane to tracer ratio.
4. Lack of TFR measurements prior to Dec 21, 2015
For the TFR method to work properly, the tracer and plume need to be well-mixed. The
methane plume exited well SS-25 with considerable upward velocity in to complex
terrain and wind patterns. These conditions create the potential for spatial separation of
leaked methane and the tracer released near the source.
The assumption of a single point source at the wellhead is simplistic for the Aliso
Canyon leak. Initially, the leak was described as coming from the entire surrounding
hillside, making the leak more of an area source. After the November 6 well kill attempt,
the crater around the wellhead formed, and it was assumed from then on all the gas
leaked from this crater. However, several NASA-JPL overflights indicate methane was
still leaking from the hillside below the wellhead to the west and east. After SS-25 had
been controlled surveys with mobile platforms conducted by ARB and the South Coast
Air Quality Management District found methane continuing to be emitted from the
surrounding hillsides. The fraction of emissions during the leak that was emitted from
the hillside vs near the well itself will likely remain unknown, but any emissions from the
hillside would be missed by the TFR.
Multiple leakage points in combination with the complex terrain and wind patterns
potentially resulted in complex plume behavior, degrading the ability to completely
measure the plume downwind and downhill from the leak site at the ground level to
establish the tracer to methane ratios as required by the TFR. The flights by Scientific
Aviation suggest that the downwind plume was often bi-modal, splitting it in two as it
was blown downwind, and often only one of the modes drifted over Sesnon Boulevard
where the TFR measurements were made (see Figure 4). An example of this is
presented in Figure 5 and complicates the determination of the tracer to methane ratios
needed to calculate the emissions.

Page 8
 

 CALIFORNIA AIR RESOURCES BOARD 
 

The uncertainty provided by SoCalGas with the TFR estimate appears to be based on
observed variability in repeated measurements, and is therefore representative of the
uncertainty in the method itself. The added steps needed to convert the measured
emissions rate to an overall emissions estimate inject additional uncertainty that should
be accounted for. This includes the lack of TFR measurements prior to December 21
2015 and the assumption that the TFR to flight measurement ratios would have been
the same during the early period of the leak.
The uncertainty in the mixing of the tracer combined with methane leakage through the
surrounding hillside, the added error associated with the assumptions on leak rates prior
to December 21, 2015, and the inability to properly capture the downwind plume from
Sesnon Boulevard, likely make the TFR estimate a lower bound and suggests the
uncertainty was greater than what has been provided by SoCalGas.

Methane Emissions (10,000 kg per hour)

8

Scientific Aviation

7

Aerodyne/TFR
6
5
4
3
2
1
0
11/2

11/12

11/22

12/2

12/12

12/22

1/1

1/11

1/21

1/31

2/10

Figure 3: Relationship between SCG/Aerodyne's TFR and Scientific Aviation's emission estimates

Page 9
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Figure 4: The terrain at Aliso Canyon is very complex leading to highly variable winds and non-linear
plume propagation patterns. The downwind measurements for the TFR method took place along Sesnon
Boulevard between the two red points on the map. The plume may not always have solely passed over
this stretch of Sesnon Boulevard.

Page 10
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Figure 5: Example of bi-modal plume measured downwind with a small airplane (top panel). The bottom
panel shows the longitudinal profile (seen as the two areas of high methane concentrations) with the eastwest span of Sesnon Boulevard. The TFR method would only have measured the eastern mode.

Page 11
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Estimates by Research Organizations
During the leak incident, there was considerable coordination between State agencies
and scientific experts to monitor, measure, and quantify the size of the leak based on
ambient measurements and other monitoring efforts. In addition to an aircraft and
several passenger cars equipped with methane monitors, the research teams leveraged
existing monitoring resources that are in place to make routine measurement of carbon
dioxide, methane, and other greenhouse gases in the Los Angeles basin. These routine
measurements are made by ARB’s statewide greenhouse gas monitoring network and a
wider collaboration with other State agencies and the federal government through the
Megacities Carbon Project, and have been effectively used in the past to inform the
State’s greenhouse gas inventories, confirm that emission reductions are occurring, and
improve the understanding of sources and mitigation options. Additional novel tools
included a satellite operated by the Japanese Space Agency, a high-altitude aircraft,
advanced remote sensing devices located on Mount Wilson and in Pasadena, and flask
measurements.
A short summary of the analytical methods to inform the quantification of the leak is
shown in Table 2 and discussed below. Collectively, these efforts have provided a
diverse data set that is being used in this report.
Table 2: Overview of other methods
Method

Project
Teams

Airborne
Measurements

Scientific
Aviation

Ground-Based
Remote
Sensing

NASA JPL,
Caltech
1.tTCCON,
2. CLARS
NASA, JPL,
NIST
[AVIRIS,
HyTES]
Hyperion
EO1,
GOSAT

Airborne
Remote
Sensing
Satellite
Remote
Sensing
Tower-Based
Measurements
and Large
Eddy
Simulation
Modeling

Megacities
Carbon
Project

Measurement Approach

Measurement
Frequency

Small airplanes with instruments measuring
methane and other pollutants flying downwind of
the leak through the plume at various elevations.
Ground based spectrometers use sunlight and its
attenuation to estimate GHG concentration in the
atmosphere above LA.

11 measurements
campaigns made
during the leak
Continuous,
operating pre-leak
and afterwards

Imaging spectrometers deployed on aircraft
image the methane plume at the leak site with
high resolution (meters) and spatial accuracy.

Only a few hours on
select days

Satellite sensors measure enhancement of
methane at different locations in and around the
LA basin. No emission estimate expected at this
point.
GHG tower network, with atmospheric transfer
models and high-resolution flow-resolving
models to directly simulate the methane plume at
the leak site and its propagation into the
atmosphere. Fitting of results to observations can
be used to estimate leakage flux.

Continuous,
intermittent,
repeated every few
days, large pixels
Continuous
measurements,
operating pre-leak
and afterwards;
intermittent plume
identification

Page 12
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Below is a short summary of the specific efforts that have resulted in emission
quantifications with the methods from Table 2.




Airborne Measurements: Steve Conley, the primary investigator for Scientific
Aviation, published another estimate based on the same overflights as used for
the ARB estimate1. Conley used a slightly different integration method than the
one used by ARB in the initial estimation, and arrived at a marginally higher
emission estimate of 97,100 MT CH4.
Ground-Based Remote Sensing: A joint research team from JPL and Caltech
operates the Total Carbon Column Observing Network (TCCON), which is a
remote sensing technique that determines the total amount of carbon in the air
column above the instrument stationed at Caltech in Pasadena. This instrument
has been in use for several years, operates continually, and was able to observe
the Aliso Canyon leak whenever any part of the plume reached the Caltech
campus. The TCCON team published the estimate for ethane (C2H6) emissions
from the leak based on measurements of LA basin-wide enhancement11.
Although the researchers did not present a methane estimate in the paper, ARB
staff utilized the ethane data to estimate the methane emissions by using
observed methane to ethane ratios. This approach yielded a total methane
emission estimate of 100,050 ± 23,562 MT CH4 Uncertainties in the ratio
estimates were compiled from different estimates of this ratio and propagated
into the methane error for the TCCON result. The error for the TCCON estimate
for the published C2H6 emission is large and the resulting error for the methane
emission total is greater than 20 percent. Since the TCCON represents a basinwide estimate and is not specifically sensitive to the Aliso Canyon leak, any
uncertainty in the basin-wide enhancements observed by the sensor will lead to
increased uncertainty in the emission estimate for the Aliso Canyon leak.

In addition to the above approaches, researchers are currently working on additional
analyses, which are expected to produce emission estimates or inform the Aliso Canyon
natural gas leak incident, as described below.


Ground-Based Remote Sensing: NASA-JPL also operates the California
Laboratory for Atmospheric Remote Sensing (CLARS) at Mount Wilson. This
remote sensing instrument is able to measure the amount of methane in the Los
Angeles basin and has been in continuous use for several years. This instrument
is able to observe the increase in the basin-wide methane due to the leak.

                                                            
11

Wunch et al.(2016), Quantifying the Loss of Processed Natural Gas Within California's South Coast Air
Basin Using Long-term Measurements of Ethane and Methane, Atmos. Chem. Phys. Discuss.,
doi:10.5194/acp-2016-359

Page 13
 

 CALIFORNIA AIR RESOURCES BOARD 
 







Emission estimates from the JPL/CLARS team using JPL/Megacities team’s
inverse modeling effort are still in progress. Preliminary findings by the inverse
modeling group indicate that the Aliso Canyon leak roughly doubled the LA Basin
methane emissions, which is broadly consistent with other measurements.
Airborne Remote Sensing: A research team from NASA-JPL and the National
Institute of Standards and Technology (NIST) is working on analysis of data from
two airborne NASA-JPL imagers (AVIRIS and HyTES) that are capable of
detecting methane. These imagers flew over the leak and resulting plume
repeatedly in January and February. The investigators are combining the
resulting measurements of methane enhancements in the plume with a highresolution numerical model (Large Eddy Simulation or LES with a flow resolution
of 10’s of meters) to arrive at independent emission estimates for the time of
overflight. This approach will not yield an estimate of total emission for the leak
but will rather serve as validation for the Scientific Aviation measurements during
Phase II of the leak.
Satellite Remote Sensing: Satellite-based sensors used by Hyperion/EO1 and
GOSAT were able to identify methane enhancements in the region, suggesting
the presence of a methane plume from space12. This method is not expected to
produce emission estimates from the leak incident, but does provide an overall
visual confirmation of the presence and flow pattern of the escaping natural gas.
Tower-Based Measurements and Large Eddy Simulation Modeling: The
Megacities Carbon Project team is currently undertaking a set of modeling
efforts, which includes inverse modeling as well as Large Eddy Simulation
Modeling efforts using monitoring data from GHG tower networks, in combination
with advanced transport models to directly resolve the methane plume at the leak
site and its propagation into the atmosphere and surrounding communities.
Although these methods are only intermittently able to identify the plume on days
when the meteorological conditions are conducive to transporting the plume from
the leak site to the neighboring monitoring stations, it nevertheless provides a
more comprehensive and continuous view of the Aliso Canyon facility from the
earliest phases of the leak. The researchers are also utilizing data from
concurrent airborne remote sensing campaigns (AVIRIS/ HyTES) to inform the
modeling analysis.

                                                            
12

Thompson, et al. (2016), Space-based remote imaging spectroscopy of the Aliso Canyon CH4
superemitter, Geophys. Res. Lett., 43, 6571–6578, doi: 10.1002/2016GL069079

Page 14
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Refinement of the Preliminary ARB Estimate
ARB evaluated the information, assumptions, and data used in the different estimation
efforts to inform the refinement of the ARB preliminary estimate of total leaked methane
emissions from Aliso Canyon. ARB also evaluated all available information from the
leak site, including pressure data from the well and the reservoir, and notes from staff at
the California Department of Conservation’s Division of Oil, Gas, and Geothermal
Resources (DOGGR), to inform our estimation approach. In the initial estimate
produced by ARB, the emission rates between the flights were assumed to be constant.
However, little was known at the time about how the leak rate varied between these
flights. Using information now available, such as a better understanding of the flow of
natural gas from the reservoir to the atmosphere, and the well pressure logs, it is
possible to adjust the assumed emission rate between the flights and update the
emissions estimate.
Well Configuration
Figure 6 shows a conceptual schematic of the leaking well, consisting of the inner well
tubing, and the production and surface casings. Early well tests suggest the leak was
caused by a breach in the 7-inch production casing at approximately 440 feet. The
cause of the leak is the subject of a root-cause-analysis still under way, and will remain
unknown until that investigation is complete. Nevertheless, the main factors that would
affect the leak rate include the pressure in the reservoir (field pressure), and the
configuration of the well, the size of the production casing breach, and the flow pathway
through the geologic media from the base of the surface casing to the ground surface.

Page 15
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Figure 6: Conceptual schematic of the leaking well (SS25)
(image provided courtesy of Lawrence Berkeley National Laboratory)

Well Control Attempts
A well control attempt in the context of this report refers to attempts to place a column of
heavy fluid into the well in order to interrupt the flow of reservoir gas through the well.
This involves pumping “heavy-weight drilling mud” or “kill fluid” into the well to establish
a large enough column of heavy fluid in the well that its hydrostatic pressure is sufficient
to counteract the field pressure. This fluid can be pumped in through the wellhead
(“top kill”), as was the case for all attempts at SS-25 except the final operation, or into
the base of the well via an intersecting well built for the purpose (“relief well”), as was
the case for the final operation. The operators can choose from different fluid densities
to achieve a hydrostatic pressure at the base of the well deemed suitable for its
geometry and the reservoir pressure.
All seven top well control attempts made at SS-25 were unsuccessful and failed to
establish sufficient downward pressure from a large enough column of “kill fluid”.
Instead, the kill fluid was observed exiting the ground surface around the well, like the
escaping natural gas. Observations suggest the ejection of the kill fluid from the well
altered the pathway through the geologic media during some of the kill attempts. In
Page 16
 

 CALIFORNIA AIR RESOURCES BOARD 
 

particular, the November 6 and 13 , 2015, well kill attempts are believed to have created
an almost unrestricted path from the bottom of the surface casing to the crater that
formed near the well head during these attempts. Subsequently, except for the field
pressure, most variables important to predicting the leak rate appear to have remained
mostly unchanged. The well was finally controlled by a kill through a relief well, which
took months to construct.
Updating the Preliminary ARB Estimate
To improve upon the simplifying assumptions made in the initial rough estimate, ARB
combined available data to conduct a detailed analysis of the emission rates from the
Aliso Canyon leak. Through conversations with staff at DOGGR and researchers at
NASA-JPL and Lawrence Berkeley National Laboratory (LBNL), ARB identified two
major phases in the course of the leak requiring separate methods to calculate the leak
rate:
1. The first phase runs from the beginning of the leak on October 23, 2015 until
November 28, 2015, and is characterized by heightened well activity such as
adding pressure gauges, installing and removing vertical pipes, installation of
equipment to stabilize the well and reduce oil misting, multiple well control
attempts and resultant changes in the geologic media surrounding the well.
These activities likely resulted in variability in emission rates during this time.
During this initial phase, wellhead pressures were available from pressure
gauges installed on the well on October 30, 2015.
2. The second phase runs from November 29, 2015 until the well was successfully
controlled on February 11, 2016. This period is characterized by less site activity,
as the focus had shifted from trying to stop the leak by a top kill to doing so from
a relief well under construction. Well pressure data during this phase is
unfortunately not available as the gauges were accidentally knocked off at the
end of November, and were not reinstalled afterwards. However, the leak rate is
expected to be steadier during this time, and driven mostly by the pressure in the
natural gas storage field. In addition, Scientific Aviation conducted regular
downwind flights during this period, and captured the leak rates with good
temporal coverage.
The following section details the relevant leak characteristics, as well as the emission
estimation approaches used in the two phases.

Page 17
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Phase I: Initial Leak and Well Control Attempts (October 23 – November 28, 2015)
Well Control Attempts:
Downwind flights:

October 24, November 6, 13, 18, 24, 25; 2015
November 7, 2015
46,900 ± 10,400 kg/hr CH4
November 10, 2015
60,400 ± 13,000 kg/hr CH4
November 28, 2015
54,700 ± 7,000 kg/hr CH4

During Phase I, there was considerable activity at the well site including six top kill well
control attempts, adding and removing lateral piping, adding pressure gauges, and
installing equipment to stabilize the well and reduce oil misting. During this period, the
State agencies were able to coordinate only three flights to measure the methane
emissions from the leak. The first such measurement occurred two weeks into the leak,
making the initial leak rate largely unknown. The scarcity of measurements, combined
with activity at the well which allows for the possibility of highly variable emissions rates,
could make simple interpolations between the airborne measurements a poor
representation of the emissions.
Pressure data indicate some of the top well control attempts reduced resistance to
leakage along the pathway through the geologic media. Specifically, the well control
attempt on November 6 is described by DOGGR staff as “re-agitating’ the leak and the
November 13 attempt resulting in a crater forming around the well and kill fluid observed
being ejected from it. The pressure in the well also dropped significantly indicating the
creation of a path through the geologic media that provided almost no resistance to
flow.
Figure 7 shows the production casing pressure data that is available from October 30 to
November 25, 2015. The reduction of this pressure on November 6 is consistent with
less resistance along the path through the geologic media. There is no data indicating
subsequent well kill attempts altered the flow pathway either in the geologic media or
the well in a manner that changed the emission rate. Unfortunately, the pressure
gauges that were installed on October 30 were knocked off by accident during the
November 25 well control attempt, and were not reinstalled for the remainder of the
leak. Pressures from October 23-30 and November 26-28 have been inferred based on
the trends in pressures during those periods to create a continuous pressure log for
Phase I.

Page 18
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Pressure [psig]

800
700

Measured SS‐25 Production
Casing Pressure

600

Extrapolated pressures

500
400
300
200
100
0
10/17/2015

10/31/2015

11/14/2015

11/28/2015

Figure 7: SS-25 Average daily production casing pressure. Pressure gauges were installed on 10/30, and
accidentally knocked off on 11/25. Pressures before and after that period are extrapolations.

Using pressure differentials between points in a pipe is often used to estimate fluid flow
and many equations have been developed to do so. Each equation is appropriate to a
different set of conditions such as type of pipe, type of fluid, and prevailing conditions
such as pressures and temperature. ARB consulted with researchers from LBNL and
selected an approach to estimate the flow in SS-25 which integrates one such equation,
the Weymouth equation, the pressure in the storage reservoir, as represented by a
pressure data from a well near SS-25, and the pressure data from production casing,
along with the rates from the downwind flights to estimate flow rates during this period.
Weymouth equation
The Weymouth Equation, presented below, is used for high-Reynolds-number flows
where the Moody friction factor is merely a function of relative roughness.
/

 

 1.1 

.

 

Where:
Q = gas-flow rate, MMscf/D,
d = pipe inside diameter, in.,
Page 19
 

 CALIFORNIA AIR RESOURCES BOARD 
 
P1 = upstream pressure, psia,
P2 = downstream pressure, psia,
L = length, ft,
T1 = temperature of gas at inlet, °R,
S = specific gravity of gas, and
Z = compressibility factor for gas, dimensionless.

In the current application the Weymouth Equation can be rearranged as follows.
 

 

 

/

Where:
Q = gas flow rate
C = a constant that includes the fluid and pipe parameters
(specific gravity, compressibility, length, diameter, and temperature)
P1 = the pressure at the upstream end of the pipe
P2 = the pressure at the downstream end of the pipe
The pressure at the bottom of SS-25, P1, is assumed to be the pressure in the natural
gas storage reservoir measured in a nearby well, and the pressure at the top of the well,
P2, is the production casing pressure at the breach depicted in Figure 7. Q in the
Weymouth equation is the gas flow rate in the well, and therefore, the leak rate.
Rather than calculating C based on well and fluid characteristics, we use the three
downwind flight measurements to find the C value that best fits all three measurements.
The method draws on both the available pressure data and the Scientific Aviation flights
to allow a calculation of emission rate for the entirety of Phase I. By minimizing the sum
of squares of the difference between the calculated Q and the estimate from the flight
results, we arrived at a C value of 23.7, which results in a total methane emission
estimate from Phase I of 48,450 MT. Figure 8 shows the results, and Table 3 shows the
calculation of the total emission during Phase I.

Page 20
 

 CALIFORNIA AIR RESOURCES BOARD 
 
80,000

Emission Rate [KgCH4/hr]

70,000
60,000
50,000
40,000
30,000
Weymouth Equation

20,000

Scientific Aviation
10,000
10/23

10/30

11/6

11/13

11/20

11/27

Figure 8: Emission rate during Phase I calculated using the Weymouth Equation, C=23.7, and the
downwind measurements by Scientific Aviation and their uncertainties.

Page 21
 

 CALIFORNIA AIR RESOURCES BOARD 
 
Table 3: Offset well and SS-25 production casing pressures and emissions calculations
Offset Well
Pressure (P1)

Offset
Observation
well

[PSIG]

Average Daily
Production Casing
Wellhead Pressure in
SS25 (P2)

Calculated Emissions
Rate

Daily Emissions

[PSIG]

[KgCH4/hr]

[MT CH4]

10/23/2015

2655

SS 5

750**

60,361

1,449

10/24/2015

2588

SS 5

725**

58,880

1,413

10/25/2015

2521

SS 5

700**

57,398

1,378

10/26/2015

2484

SS 5

675**

56,656

1,360

10/27/2015

2447

SS 5

650**

55,910

1,342

10/28/2015

2436

SS 5

625**

55,801

1,339

10/29/2015

2424

SS 5

600**

55,661

1,336

10/30/2015

2416

SS 5

582

55,575

1,334

10/31/2015

2407

SS 5

585

55,336

1,328

11/1/2015

2397

SS 5

672

54,532

1,309

11/2/2015

2387

SS 5

630

54,565

1,310

11/3/2015

2379

SS 5

620

54,432

1,306

11/4/2015

2371

SS 5

524

54,802

1,315

11/5/2015

2364

SS 5

552

54,477

1,307

11/6/2015

2358

SS 5

351

55,261

1,326

11/7/2015

2355

SS 5

214

55,583

1,334

11/8/2015

2354

SS 5

204

55,579

1,334

11/9/2015

2352

SS 5

211

55,517

1,332

11/10/2015

2350

SS 5

214

55,463

1,331

11/11/2015

2340*

SS 5

229

55,189

1,325

11/12/2015

2330*

SS 5

240

54,920

1,318

11/13/2015

2319*

SS 5

217

54,730

1,314

11/14/2015

2309*

SS 5

229

54,460

1,307

11/15/2015

2299*

SS 5

223

54,232

1,302

11/16/2015

2289*

SS 5

211

54,016

1,296

11/17/2015

2279*

SS 5

208

53,782

1,291

11/18/2015

2269*

SS 5

219

53,515

1,284

11/19/2015

2258*

SS 5

209

53,294

1,279

11/20/2015

2248*

SS 5

203

53,065

1,274

11/21/2015

2238*

SS 5

201

52,828

1,268

11/22/2015

2228*

SS 5

199

52,589

1,262

11/23/2015

2218*

SS 5

199

52,348

1,256

11/24/2015

2208*

SS 5

194

52,115

1,251

11/25/2015

2197*

SS 5

202

51,857

1,245

11/26/2015

2187*

SS 5

200**

51,619

1,239

11/27/2015

2177

SS 24

200**

51,377

1,233

11/28/2015

2153

SS 24

200**

50,805

1,219

Total Emissions for
Phase I →

48,445

* November 11-26 Pressures have been adjusted to account for withdrawals
from SS 5
** October 23-29 and November 26-28 pressures are extrapolations

Page 22
 

 CALIFORNIA AIR RESOURCES BOARD 
 

For Phase I, the uncertainty of the emission rate for each day was calculated using error
propagation for the Weymouth Equation, starting with errors for each parameter in the
equation: the coefficient C has an uncertainty of ± 4.3, the maximum amount of
variability in C that still puts the flow rate calculated within the uncertainty bounds of the
three downwind flights, while P1 was assigned a 1 percent uncertainty and P2 was
assigned a 6 percent uncertainty.
The 1 percent uncertainty value for P1 is approximately ten times the variation from
November 7 to November 10, a period of relative stability that provides some
perspective on the measurement uncertainty. The larger value was chosen to account
for the possibility of some nonlinear variation between the pressures at the offset well
and the base of SS-25 for which it was used as a proxy. P2 was assigned a 6 percent
uncertainty based upon the range of the central 95 percent of production casing
wellhead measurements from 9:53 to 22:41 on November 14. This period was selected
because the overall trend was nearly zero (no change in pressure).
These daily errors were added to give a total uncertainty for the estimate of methane
leaked during Phase I of ± 8,810 MT CH4.
Estimate of Methane leaked during Phase I: 48,450 ± 8,810 MT

Page 23
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Phase II: Drawdown (November 29, 2015 – February 11, 2016)
Kill Attempts:
Downwind flights:

December 22, 2015
December 4, 2015
December 12, 2015
December 23, 2015
January 8, 2016
January 12, 2016
January 21, 2016
January 26, 2016
February 4, 2016

48,800 ± 9,300 kg/hr CH4
38,000 ± 4,800 kg/hr CH4
28,100 ± 4,800 kg/hr CH4
22,700 ± 3,300 kg/hr CH4
20,700 ± 3,100 kg/hr CH4
19,100 ± 4,000 kg/hr CH4
19,000 ± 1,700 kg/hr CH4
22,500 ± 3,800 kg/hr CH4

During Phase II there was much less activity near the well as the focus shifted to drilling
the relief well, and the leak rate was mainly driven by the field pressure which was
steadily reduced with the withdrawal of natural gas from the reservoir. The only kill
attempt during the Phase is not expected to have affected the pathway through the
geologic media, and therefore, not expected to have affected the leak rate.
The period was well sampled with 8 downwind flights. For this portion of the leak,
although well pressure data are not available, non-linear changes in leak rates are not
expected since there were no major well control attempts. Therefore, we used simple
interpolation of the measured leak rates from the downwind flights between the
sampling dates. The uncertainty for Phase 2 was calculated based on the uncertainties
provided by Scientific Aviation for the downwind flights as listed above. The error was
propagated by taking the square root of the sum of squares of the errors for each
specific period when summing over all periods for the total emission for Phase 2.
Estimate of Methane leaked during Phase II: 51,200 ± 2,970 MT

Page 24
 

 CALIFORNIA AIR RESOURCES BOARD 
Table 4: Methane Emission Estimates of the Aliso Canyon Natural Gas Leak
Phase
Phase I
Phase II

↓

Flight Date

11/28/2015
12/4/2015
12/12/2015
12/23/2015
1/8/2016
1/12/2016
1/21/2016
1/26/2016
2/4/2016

Leak Rate                    
Uncertainty                   Assumed number of 
[kilogram methane per hour] [kilogram methane per hour] days at this leak rate
54,700
48,800
38,000
28,100
22,700
20,700
19,100
19,000
22,500

*   Assumes natural gas from the leak is 94% methane, and methane and NG has density of 0.01858 kg/cu‐ft

7,000
9,300
4,800
4,800
3,300
3,100
4,000
1,700
3,800

3.0
7.0
9.5
13.5
10.0
6.5
7.0
7.0
11.5
Total

Leaked methane for this period                                            
[kilogram methane]                                [billion cubic feet of natural gas, bcf*]
48,444,715                                                                      2.8
                                             3,938,400                                                                      0.2
                                             8,198,400                                                                      0.5
                                             8,664,000                                                                      0.5
                                             9,104,400                                                                      0.5
                                             5,448,000                                                                      0.3
                                             3,229,200                                                                      0.2
                                             3,208,800                                                                      0.2
                                             3,192,000                                                                      0.2
                                             6,210,000                                                                      0.4
                                          99,637,915                                                                      5.7

Page 25 
 

 CALIFORNIA AIR RESOURCES BOARD 
 

Total Emissions: Phase I + II (October 23, 2015 – February 11, 2016)
Finally, the total estimated methane leaked during both Phases was calculated by aggregating the emissions estimated in the
two phases (Figure 9 and Table 4). The combined error was calculated by taking the square root of the sum of squares of the
individual Phases.
80,000
Estimated Emission Rate

Leak Rate [Kg CH4/hr]

70,000

Scientific Aviation

60,000
50,000
40,000
30,000
20,000
10,000
10/23

11/6

11/20

12/4

12/18

1/1

1/15

1/29

2/12

Figure 9: The leak rate over the duration of the incident and the downwind measurements.

Estimate of Methane leaked during Phases I - II

= 48,450 ± 8,810 MT + 51,200 ± 2,970 MT
= 99,650 ± 9,300 MT

Page 26
 

 CALIFORNIA AIR RESOURCES BOARD 

Final emission estimates and amount of methane that needs to be
mitigated
As described in the above sections, ARB calculated the updated emission estimate of
99,650 ± 9,300 MT methane based on the best available data on the leak, including
pressure data from well SS-25 and an offset well, as well as direct measurements of
emissions rates made with the small instrumented airplane that flew through the
methane plume.
ARB also reviewed the revised estimate in relation to all the other available estimates
from SoCalGas, DOGGR, LBNL, and NASA-JPL to corroborate the data and inform the
final number. ARB utilized the uncertainty range defined by expected errors in each of
the numbers and calculations in order to conduct an equivalent comparison (Figure 10).
The comparison chart shows that updated number is within the error bounds of all other
estimates published, except for the SoCalGas tracer flux method.
 
180,000
160,000

Total Methane Emissions (MT CH4)

140,000
120,000
100,000
80,000
60,000
40,000
20,000
0
ARB Initial SCG Inventory SCG Tracer
Estimate

Conley

TCCON

ARB Updated
Estimate

Figure 10: Overview of different emission estimates

Page 27 
 

 CALIFORNIA AIR RESOURCES BOARD 
 

The Governor’s Proclamation directs ARB to provide a plan that fully mitigates the
climate impacts of the methane leaked from the Aliso Canyon Natural Gas Storage
Facility. The updated emissions estimate represents the most likely result with an
expected uncertainty. The final number must recognize that “fully mitigating” the
environmental damage, as required by the Governor’s Proclamation, requires taking a
conservative, or upper bound, estimate to mitigate the full potential emissions from the
leak. Therefore, to ensure that the Aliso Canyon natural gas leak is fully mitigated, the
total amount of methane that needs to be mitigated is 109,000 MT.

Page 28