Are California’s cogeneration systems performing up to par?

By Lyn Corum

A report completed by Itron Inc. in February 2007 found that cogeneration systems partially funded by the State of California’s Self-Generation Incentive program (SGIP), designed to promote distributed generation, were not performing to state standards.

Flawed design assumptions and equipment failures were at the heart of the problems plaguing cogeneration systems powered by both microturbines and internal combustion engines.

The report recommended that cogeneration systems be designed to minimum electrical and thermal loads, that actual electrical values be used in calculations, and that load profiles and hours of operation be better developed for the application process.

Itron was asked by the SGIP Working Group in December 2005 to do the analysis. But larger events caught up with the program before the recommendations could have any impact on or improve cogeneration system efficiencies.

Both Sara Birmingham, manager of the SGIP program at Pacific Gas and Electric Co., and Jon Bonk-Vasko, SGIP program manager at the California Center for Sustainable Energy in San Diego, are members of the SGIP Working Group and are familiar with Itron’s report. (The working group is made up of representatives from three of the state’s major investor-owned utilities, who hand out the incentive dollars, and the California Energy Commission. The California Center for Sustainable Energy manages its share of the program for San Diego Gas & Electric.)

Birmingham and Bonk-Vasko explained that new California Air Resources Board standards have severely affected the ability of microturbines and internal combustion units to participate in the program. As a result, applications for SGIP incentives for fossil fuel-fired cogeneration systems have dried up.

Furthermore, a 2006 state law (AB 2778) restricted technologies eligible for funding to wind and fuel cells, starting in 2008.

Solar technologies were removed from the SGIP program and given their own funding under the California Solar Initiative.

Therefore, said Bonk-Vasko, redeveloping the waste-heat utilization metrics on the application as recommended by Itron would have no impact on the cogeneration projects applying for incentive payments in 2007.

And the costs associated with any application modifications are such that making the changes are not a priority. Birmingham agrees, saying, “It doesn’t seem prudent to act on any revisions because of the limited market.”

Birmingham says that the application—known as the waste heat utilization worksheet—has indeed been modified to make it more straightforward this year, but Itron’s recommended changes that would improve efficiencies were not incorporated. However, if cogeneration systems do become eligible for SGIP funds in the future, “I think it would be prudent to look at Itron’s recommendations and discuss whether the standards are appropriate,” she says. “I would love to have a workshop with the cogeneration community to look at the efficiency standards.” The CPUC adopted standards for the SGIP program, which were created in 1978 as part of the US Public Utilities Regulatory Policy Act (PURPA). “Should they be revisited?” she asks.

David Bruder, manager of non-residential energy-efficiency programs at Southern California Edison, agrees that the Itron report has been overcome by events and the CPUC’s policy direction, which is now emphasizing renewables. “Our feeling is that with the California Solar Initiative and the remaining incentives for biogas fuel generation, it fits with our objective of a better environment,” he says.

Bruder also agrees with Birmingham and Bonk-Vasko. “We could make [the worksheet] more complex and rigorous including pre-installation data collection of thermal loads, but it will add costs with the diminishing rate of systems being installed,” Bruder says.

Keith Davidson, president of DE Solutions, is very familiar with SGIP. “There is truth to it,” he says, referring to the Itron report.

Most of the systems Itron looked at were installed five or six years ago, he says. Some of the developers were able to make the economics work without documenting how well heat could be used, so a lot of systems were oversized, Davidson explains.

As time goes on, gas prices go up, Davidson continues, and in order to make the economics work on day one, you have to have good heat utilization. Cogeneration systems should not be operating when their heat loads don’t justify it. This means sizing to the minimum coincident load, and the system should be load following.

As for the application process, Davidson suggests the SGIP program could provide more helpful information. For example, typical applicant profile data might be made available on the Web site to demonstrate that applicants must think through load profiles.

Incentives for gas-fired distributed generation systems will end on December 31, 2008, when all cogeneration systems seeking incentives have to be online.

Davidson is one of several stakeholders attempting to get new state legislation to extend that deadline. An assembly bill already drafted, AB 1064, has been put off for action until 2008.

A Little History
The California Public Utilities Commission has provided funding to investor-owned utilities since 2001 to provide incentives for their customers to install distributed generation. Technologies that are funded include solar and wind turbine systems, fuel cells, microturbines, and internal combustion engines. The SGIP requires that systems burning fossil fuels have to adhere to efficiency standards patterned after those found in PURPA.

The CPUC adopted the following waste heat recovery and efficiency standards for nonrenewable facilities obtaining SGIP incentives: At least 5% of the facility’s total annual energy output must be in the form of useful thermal energy, and the useful annual power output plus one-half the useful annual thermal energy output equals not less than 42.5% of any input of natural-gas or oil energy.

By the end of 2006, 187 MW had been installed as a result of SGIP program incentives. Cogeneration systems represented 53% of the installed capacity, although they got 14% of the funds through 2005. Installed solar photovoltaic (PV) systems represented 40% of installed capacity and got 68% of the funding, also through 2005.

Current incentives for microturbines and internal combustion engines fired by renewables are higher ($1.30 and $1.00 per watt, respectively) than those for fossil-fueled systems ($0.80 and $0.60 per Watt, respectively). In 2005, solar PV installations received an incentive payment of $3.50 per Watt, which explains why they got the lion’s share of the funding.

In December 2005, the SGIP working group asked Itron to investigate whether cogeneration facilities were achieving the required levels of energy efficiency and useful waste heat recovery. Its final report, dated February 2007, found that both internal combustion engines and microturbines surpassed by far the useful thermal requirement of 5%, achieving an average performance of 43.4% thermal energy efficiency. However, more than two-thirds of the systems failed to meet the plant efficiency goal of 42.5%. The average performance was 36.8% efficiency.

Itron has been conducting periodic measurement and evaluation of distributed generation facilities installed under the program for the SGIP Working Group. Its access to directly metered data, including amount of electricity generated, heat recovered, and fuel used at each site, allowed it to estimate actual achieved efficiencies and useful waste heat recovery levels for each facility. It was able to identify those facilities that had significant problems complying with the requirements and those that exceeded the requirements. In-depth analyses, surveys, and site visits were then used to develop recommendations to improve efficiencies and better recover useful waste heat.

At the end of 2005, there were 129 microturbine applications in the SGIP, but only 81 were operational and 71 operated on natural gas. The remainder operated on renewable or mixed fuels. Only 19 microturbines were metered for heat recovery. All projects operating on renewables are not subject to heat recovery requirements and are not required to be metered for heat recovery.

There were 232 internal combustion engine systems in the program as of December 2005, and of the 126 systems operating, 122 are fueled by natural gas, and 33 were metered for heat recovery.

Itron analyzed 47 sites for system efficiency. These included 28 internal combustion engines, 18 microturbines, and one fuel cell. Out of these, only 12 sites are good performers, having reached or bypassed the CPUC’s required efficiency level of 42.5%. The remaining 35 had efficiencies under 42.5%. The specific findings are reviewed below.

Useful Waste Heat Recovery
Microturbine installations ranged in size from 28 kW to 600 kW. Another was 1,383 kW. All exhibited a wide range of useful heat recovery, from 22% to 71%, easily meeting the 5% requirement for recovered heat efficiency.

There is a slight improvement in efficiency with larger systems, but this was attributed to the additional design-stage engineering involved in the larger projects.

Itron explored why some systems had lower waste heat recovery than others. It discovered that many of these systems’ calculations used at the design stage had significant flaws in assumptions that overstated achievable efficiency.

One example of a bad assumption that led to significant impacts on efficiencies was the overstating of hours of operation of the proposed system. Another bad assumption was that of not quantifying the coincidence of electrical demand with thermal demand. This should be done on the waste heat utilization worksheet used in the application process for state rebates, the report stated. The result was a mismatch of load and generation that possibly led to oversizing of equipment.

In addition to flawed design assumptions, equipment failures also contributed to poor performance. In some cases, whole systems shut down. At other times, only part of the system—the heat recovery loop, for example—was disabled, and the generator continued to produce electricity without recovering heat, causing efficiencies to plummet.

Some of the equipment failures included failure of the heat exchanger due to unexpected reactions with working fluids, failure of the gas compressor, or failure of the absorption chiller.

Internal Combustion Engines
The 28 internal combustion engines for which data were examined ranged from 60 kW to 1,500 kW. The study found a significant variation in the amount of useful waste heat recovery per unit of installed capacity and the efficiency of useful waste heat recovery among the systems. Efficiencies ranged from 15% to 69%. This led Itron to observe that the larger systems did not perform as well as the smaller systems sized between 60 kW and 150 kW.

The study looked at why some systems performed poorly. As with the microturbines, the calculations at the design stage often had significant flaws that overstated the achievable efficiency. Hours of operation were routinely overstated, and applications did not quantify the coincidence of electrical demand with thermal demand. This resulted in a mismatch and possibly oversizing of equipment.

Also, equipment failures contributed to poor performance. In some cases, failures caused complete system shutdowns that did not affect regulatory efficiency calculations. However, these calculations were affected when just part of the system was disabled, as in the previously mentioned case in which a heat-recovery loop fails while the generator continues to produce electricity without recovering heat, causing efficiencies to plummet.

Recommendations
Itron recommended that electric and thermal load profiles be developed as part of the application process. Even though they are key indicators of the cogeneration system’s future success, they are often estimated using assumptions and estimates that are not documented.

Furthermore, the system should be designed to serve the minimum electrical and thermal loads. Coincident electrical and thermal loads are imperative. Without them, the cogeneration system should be downsized to meet the minimum of one or the other loads. This will ensure full-load operation of the microturbine.

Electrical and Overall System Efficiency
Itron defined electrical system efficiency as electrical output divided by fuel input. Overall system efficiency is the sum of the electrical output and the useful waste heat recovery divided by the fuel input. California PUC requirements compare recovered waste heat to the total energy output from both the thermal and electricity generation contributions. This means that electrical efficiency counts twice as much as thermal efficiency.

Microturbines
Electrical system efficiencies for microturbines were significantly and consistently below manufacturer claims of approximately 30%. All 18 analyzed had electrical efficiencies between 15% and 25%. However, overall system efficiency varied between 20% and 73%.

Itron concluded that the source of the variation was the result of useful waste heat recovery—not on the electrical conversion side—and that there are operational issues on both sides of the cogeneration system. CPUC regulations require that overall system efficiency must be 42.5% but the microturbines examined achieved an average of 34%. The minimum level achieved was only 18% and the maximum level achieved was 47%.

Itron studied the design assumptions contained in the SGIP applications and found the following. Parasitic loads on the generator were not accounted for, thereby overstating electrical output. Electrical conversion efficiencies were routinely overstated, as well as hours of operation. Also, fuel might be reported in high heat values rather than low heat values. Again, coincidence of electrical demand with thermal demand was not quantified, resulting in a mismatch and possible oversizing of equipment.

Documented equipment failures that contributed to poor performance included the following: Heat exchangers failed because of unexpected reactions with working fluids. The recuperator failed causing poor electrical power output. The gas compressor or the absorption chiller failed. On projects using renewable fuels, poor fuel quality led to part-load operation. Further, operating temperatures had an effect on system electrical efficiency. Overall, regular maintenance is required to maintain good efficiency.

Internal Combustion Engines
These cogeneration systems generally had higher electrical efficiencies than microturbine systems. Itron suggested that this might be because of their relatively longer history and many of the issues discovered during product development have been resolved.

There were a wider range of electrical efficiencies (17% to 34%) and no correlation between efficiency and system size, as was the case with the microturbines. Overall system efficiencies range from 31% to 69%.

In general, internal combustion engines complied more closely with CPUC efficiency requirements than did microturbine systems but overall did not meet the requirements. The average level of efficiency was 38%, and the minimum level achieved was only 28%. The maximum level achieved was 48%.

The inability to meet the CPUC efficiency requirements was attributed to not accounting for the generator’s parasitic loads on the application, overstating electrical conversion efficiencies, overstating hours of operation, and not quantifying the coincidence of electrical demand with thermal demand.

Equipment failures that contributed to poor performance included heat exchanger failures due to unexpected reactions with working fluids, recuperator failure causing poor electrical power output, gas compressor failures, or absorption chiller failures. Other equipment failures were due to poor quality of renewable fuel (leading to part-load operation) and operating temperatures affecting electrical efficiency.

Again, Itron stressed that regular maintenance is required to maintain good efficiency.

Itron’s Recommendations
The SGIP application should require actual electrical efficiency instead of entering nameplate values that do not adequately account for parasitic loads or part-load operation.

Documentation illustrating electric and thermal load profiles should be required. Again, applicants often estimated electric and thermal loads without
documentation. The system should be designed to minimum electrical and thermal loads. If there is no coincident electrical and thermal load, the cogeneration system should be downsized to meet the minimum of either load. This will ensure full-load operation of the microturbine or internal combustion engine.

Manufacturer Efficiencies
Capstone and Ingersoll Rand systems dominated the microturbine applications—they represented over 70% of installed capacity. Waukesha, Hess Microgen, and GE Jenbacher systems represented over 48% of the internal combustion engine cogeneration applications.

There were only marginal differences between the efficiencies of Capstone and Ingersoll Rand microturbine performances when compared to the rest of the manufacturers’ turbine performances. Capstone had marginally lower efficiencies that Itron attributed to a large number of system components, such as heat exchangers and controls. These drew a parasitic load from the electrical generation that may not have been captured in the monitoring data of the non-Capstone microturbines.

Second, early generation Capstone units had equipment problems that compromised the operational performance of the units.

Most participants in the study that experienced equipment problems claimed that Capstone replaced the equipment under warranty after performance data had been collected. Future monitoring comparisons on the replaced units may produce higher efficiency results.

The comparison of internal combustion engine units among manufacturers revealed that Hess Microgen units performed better than their competitors. Itron suggested that a relatively large equipment installer/operator routinely provided service along with operation. As a result, the Hess equipment not only operated more efficiently; it operated at a higher-capacity factor.

Overall Observations
Microturbines were found to have a higher average capacity factor—43% compared to 37% for internal combustion engines. This is the percent of time the unit is available at the rated capacity over the entire data range of valid monitoring data per site.

While this is true, the amount of work available from the capacity is less. Microturbines typically have higher thermal efficiencies, but since internal combustion engines beat out microturbines on electrical efficiency, they do not need to recover as much heat to comply with the CPUC regulation for overall system efficiency. However, microturbines show a more favorable result for proportion of facilities’ total annual energy output in the form of useful heat.

Final Thoughts
Ron Ishii is vice president of Alternative Energy Systems Consulting Inc. and is under contract to provide technical analysis and other assistance to the SGIP Working Group.

Itron is correct, he says. Units running at low partial loads will have degraded electrical efficiencies. The size of the cogeneration system needs to be lowered to serve the baseload, and this will improve the overall fuel efficiency.

Ishii says the way the program design was intended, to some degree, was to create good system design. Still, it requires quite a bit of technical support for sizing to achieve overall efficiencies.

Lyn Corum specializes in energy topics.

DE - November/December 2007