Aliso Canyon Panorama

Aliso Canyon.

Several simulations of the Southern California natural gas infrastructure system indicate Aliso Canyon might still be needed to ensure system reliability and keep costs low in extreme operational conditions that occur once every decade, although the system might be able to get by without supplies from the natural gas storage facility.

The findings were presented at a California Public Utilities Commission online meeting held July 28.

The meeting is one of several in an investigation opened by the CPUC in February 2017 [I17-02-002], as outlined in SB 380 following the 2015 natural gas leak at Aliso Canyon. The bill text also specifies that the CPUC is required to work with other entities in making its determination, including the California Energy Commission; the California Independent System Operator; publicly owned utilities that rely on Aliso Canyon natural gas for electricity generation; the California Geologic Energy Management Division of the state Department of Conservation (formerly the Division of Oil, Gas, and Geothermal Resources); relevant government entities; and others.

In 2017, the CPUC set a 10-year timeline for the closure of the natural gas storage facility in Southern California; however, assessments of the state's energy demands have shown a need for natural gas generation into the foreseeable future.

Southern California Gas Co., the storage facility's owner, contends the continued operation of Aliso Canyon is necessary to maintain affordable, reliable energy in Southern California, pointing to five previous studies supporting its claim (see CEM No. 1566).

Local residents and environmental activists have been pressing for the facility's closure since a massive gas leak was detected there on Oct. 23, 2015. The leak was declared plugged on Feb. 18, 2016. A root-cause analysis report stated that the outer 7-inch well casing ruptured due to microbial corrosion and that there were more than 60 such casing leaks at Aliso Canyon prior to October 2015, stretching as far back as the 1970s. The report also stated that SoCal Gas never conducted a failure investigation, nor any follow-up inspections or analyses after previous leaks (see CEM No. 1539).

The first events in the CPUC's ongoing investigation proceeding—a workshop in Northridge and a prehearing conference in Los Angeles—were both held April 17, 2017.

The July 28 web meeting, the third in a series of workshops on the issue held to date, presented various elaborate modeling exercises conducted by both the CPUC and Los Alamos National Laboratory staffs, some of which attempted to validate information provided by SoCal Gas. It was, in essence, a "show your work" session focused on six different scenarios related to systemwide natural gas conditions in Southern California across two seasons in three different years—2020, 2025 and 2030.

The various calculations underpin what will ultimately be a decision by the CPUC about the long-term viability of the Aliso Canyon facility. This could include its curtailment or closure, depending on whether systemwide energy reliability is at risk from such a change.

"The scope of the proceeding does not include the question of whether the facility should be reopened for injections, but rather the long-term feasibility of minimizing or eliminating the use of the facility while still maintaining energy and electric reliability for the Los Angeles region, consistent with maintaining just and reasonable rates," the CPUC says on its website.

Any policy decisions are at least four months away. It's the policy decision that everyone is awaiting, but the math is the foundation on which the decisions will ultimately be based, which is why there was periodic debate during the webinar about data sources and whether a given model or approach was appropriate to use.

As Donald Brooks, supervisor of the CPUC modeling team, explained, the modeling presented at the workshop was designed to identify any gaps or needs that might be present if Aliso Canyon operations were curtailed or the facility closed. The simulations were based on current information on the system infrastructure, rules and regulations. Based on the results, changes could be warranted—but discussion on this point is the express purpose of the forthcoming Phase 3 workshop.

Although SoCal Gas did not present any data, both the CPUC and LANL staff were tasked, in part, with replicating existing modeling completed by the utility to verify its calculations. Environmental advocates have previously questioned the reliability of SoCal Gas' data modeling.

One of the first models discussed by Mounir Fellahi of the modeling team was production cost modeling. The team looked at SoCal Gas operations in which there was unconstrained generation—generation without any curtailment—and minimum local generation—conditions in which the system meets the Federal Energy Regulatory Commission's basic local reliability criteria and in which some power plant operations are curtailed.

The team found that reliability degraded in all six scenarios it studied based on the likelihood of one extreme event occurring within a 10-year period. It also found that, in 2030, the cost of operations under the minimum local generation scenario was $121.3 million greater than for the unconstrained scenario. Greenhouse gas emissions were also reduced in that year, ostensibly because available supplies were unable to serve demand.

There were several instances throughout the workshop in which some of the data points or facility names were redacted, with confidentiality being cited as the reason for the absent information.

The often-repeated answer to many questions about the numbers in the 2024-2025 time frame was the Diablo Canyon nuclear power plant. In the minimum local generation case for September 2025, for example, there is roughly 3,600 MWh of unserved load—largely attributed to the loss of generation from the nuclear plant. The numbers provided by the California Independent System Operator used in the simulation reflect retirement of the facility, CPUC staff said.

Still, many participants were concerned about the origins of the varied data and how the CPUC did its math—SoCal Gas in particular.

One person in attendance who identified as being with SoCal Gas took the commission to task for stating that there were problems in the system.

"We're not mitigating things," Brooks countered during the exchange. "We're pointing out things to be fixed."

"But why are you saying there's a problem? It's not a reliability issue. It's an investment issue," the SoCal Gas representative responded.

Brooks cut short the debate, saying, "I think I understand you and you understand me."

But it would not be the last frustration expressed in the question-and-answer sessions that peppered the online event. One issue noted several times was the limitations of available modeling tools, which aren't designed to answer the CPUC's specific questions.

As Anatoly Zlotnik, who has been working on Aliso Canyon modeling for several years through LANL, noted, "There is no commercial software that can solve this problem . . . The system is large and complicated . . . The answers are not simple . . . SoCal operators do quite a good job given the tools they have." It could be possible, for example, to eventually develop a tool to evaluate a unique issue such as zonal capacity, Zlotnik said.

Others were frustrated with assumptions made about the operation of lines 235-2 and 4000, the status of which is an ongoing sore spot (see CEM No. 1526). One speaker said the reduction in pressure on those lines "should not be part of this simulation because this is an operational issue SoCal Gas has decided to work with" rather than based on available capacity. Staff countered that they do not see the lines "coming back to capacity." Were SoCal Gas to fix the line outages, then the assumption would be that supplies would increase; however, the CPUC simulations are designed to mesh with how the pipeline has traditionally been operated, they said.

Lisa Cosby of the CPUC's energy resource modeling team delved into a single scenario to evaluate whether the system could operate without Aliso Canyon on a 1-in-10 winter day. The simulation failed. System pressures were below the minimum operating level, demand was not met, the linepack was not restored, and roughly 731 MMcfd would be needed to meet demand on that modeled winter day. Cosby reiterated when pressed that "even with the best-case scenario, Aliso Canyon is needed. That's what this simulation is saying."

In their presentation, Zlotnik and Mary Ewers of the Los Alamos lab, who conducted an independent review of the hydraulic modeling, said that looking at high-demand situations on one day in every 10 years—which reflects conditions such as a heat wave or a cold snap—is useful because not every potential system-reliability issue can be analyzed.

The key point to take from the modeling is that even when the system is stressed, it can still "get by" without Aliso Canyon, Zlotnik said. Ewers later reiterated that point when questioned.

The CPUC's Energy Division has begun accepting informal questions and comments on the presentations, according to staff. It is still working on the modeling for a 1-in-35, or extreme peak demand, scenario. Those results will be presented during a shorter webinar in the fall. The Phase 3 workshop is also forthcoming, and a firm was recently awarded a contract by the commission for work on that phase of the process.

A report containing the final results from the production cost and hydraulic modeling that will have been presented in these multiple-phase workshops will be published in fall 2020. This will be used "to help us finalize a decision on this proceeding," Brooks said.