Dr. Brad Borum, Director of the Electricity Division of the Indiana Utility Regulatory Commission (IURC) filed his draft report on the four (4) Integrated Resource Plans (IRPs) filed in 2013 by
- Duke Energy Indiana (DEI),
- Indiana Michigan Power Company (I&M),
- Indiana Municipal Power Agency (IMPA), and
- Wabash Valley Power Association (WVPA).
IndianaDG Readers may be interested in reading Dr. Borum's entire draft report which can be found at:
Of greatest interest, however, to IndianaDG Readers may be the following excerpts from the reports concerning DG and Renewable Energy. There is also a discussion of Energy Efficiency (EE) as well in Borum's Draft Report.
Indiana Michigan Power Company (I&M)
Treatment of Distributed Generation
The purpose of this section is to primarily discuss how I&M modeled distributed generation (DG) in the IRP modeling exercise, but we will also touch on some aspects of how utility scale renewable energy was analyzed.
Assumptions
I&M notes that the cost of solar panels has declined considerably over the last decade and that various forecasts generally see declining nominal prices for the next decade (p. 125). They also recognize that distributed solar, often seen on rooftops, is also experiencing declining costs as associated hardware, such as inverters, racks, and wiring bundles become standardized. The result is that both distributed and utility scale solar projects will be more economical in the future.
Utility scale solar up to 50 MW per year of incremental nameplate capacity was made available to the Plexos optimization model for selection beginning in 2014. One assumes the installed cost for solar panels in Figure 5D-3 is reflected in the costs used in the Plexos model.
Distributed solar resources were modeled at their cost to the utility which I&M stated is the full retail net metering rate, not the installed capital costs.
I&M observes that the cost of electricity from wind generation is becoming competitive within PJM due in large part to subsidies such as the federal production tax credit and REC values. Wind resources are modeled as Purchase Power Agreements with costs at constant real rate of $65 per MWh. I&M limits the implementation of wind resources to a “realistic amount,” 100 MW, each year in the Plexos modeling. An assumption made by I&M is that the Federal Production Tax Credit will not be extended beyond 2013. Distributed wind was not modeled in developing this IRP.
Biomass and incremental hydroelectric resources were not considered in the modeling process. [Emphasis added]
Results
The presentation of the results of the optimization modeling and the development of the Preferred Portfolio is confusing.
On page 184, I&M says the optimization modeling process did not select any distributed solar even though their costs decline throughout the planning period. The costs referred to appears to be the installation capital costs although this is not made clear. According to I&M the reason for this is that the solar DG resources were modeled at a cost based on the full net metering rate. I&M also presents a Figure 4E-3 on page 93, duplicated in Figure 8C-2 on page 184, which presents four different lines on a graph:
1. A line representing Net Metering Payments
2. A line representing the PJM Value of Solar
3. A line representing Utility Scale PV with the Investment Tax Credit (ITC)
4. A line representing Consumer Scale PV with ITC
I&M describes this graph as showing the avoided cost value of a typical rooftop resource in relation to its net metering cost (p. 184). On page 92, I&M says, referring to the table on page 93, customer-sited DG costs the utility more than the PJM value it provides.
The presentation of the DG solar analysis is flawed because the reader has no means to understand what I&M did. The information presented in Figures 4E-3 and 8C-2 is described in one or two sentences and provides no information as to how the data presented was developed, the sources of the data, and the assumptions required to develop the data.
Given that I&M modeled DG solar using the “full retail net metering rate,” it would have been useful to explain exactly what this rate included and how it was calculated. It would also have been instructive to perform an optimization using some different assumptions instead of only using the retail net metering rate.
As noted above, I&M developed two optimized portfolios using the base commodity forecast and two different load forecasts (Old and New). Table 8C-1 shows the summary capacity additions for the two optimization portfolios. The table indicates that 249 MW of utility scale solar is added in the period 2020-2033. The 249 MW is based on the PJM capacity value which recognizes 38% of solar nameplate MW capacity for ICAP purposes. This means I&M is projecting the addition of 700 MW of utility scale solar to be added over the period 2020-2033.
I&M then constructs a final “Preferred Portfolio” based on the portfolio optimized under the new load forecast. The Preferred Portfolio begins to add distributed solar in 2016 “at a point that roughly corresponds to the cross-over point in value from the customer’s perspective.” (p. 185) By 2033, 153 MW (nameplate) of DG solar are added on the customer side of the meter. I&M states ‘this portfolio is identical to the optimized portfolio with the addition of over 150 MW (nameplate) distributed generation through the planning period that is thought likely to occur under current net metering compensation rules.” (p. 185)
The problem is that I&M added the solar DG because it was “thought likely to occur.” So the solar resource additions appear to be ad hoc in nature and no more explanation is provided. How did I&M derive what it thought was likely to occur? [Emphasis added]
Duke Energy Indiana (DEI)
Renewables and Distributed Generation
The discussion of distributed generation in the IRP is minimal and DG is not explicitly modeled in the resource portfolio development exercise except to satisfy a minimum level of renewable generation for each scenario. Customer self-generation is discussed in two short paragraphs on page 31 in the load forecast chapter. There DEI says no additional cogeneration units that impact the load forecast are assumed to be built or operated within the DEI service territory over the forecast period. DEI goes on to say the renewables or EE categories in this IRP can be considered placeholders for any new cogeneration projects.
Non-utility generation as future resource options is discussed on page 69. DEI states a customer’s decision to self-generate or cogenerate is based on economics, and that such projects are generally uneconomic for most customers. As a result, DEI says it does not attempt to forecast specific megawatt levels of this activity. It is argued that cogeneration facilities that are built affect customer energy and demand and are captured in the load forecast. Again, DEI says that portions of the projections for renewables and EE in the IRP can be viewed as placeholders for these types of projects.
Utility scale solar is discussed at the bottom of page 74 and continuing to page 75. Screening curves are developed for 150 MW wind and 25 MW solar PV. According to DEI, solar is the least expensive but has a 20% capacity factor and has greater contribution at system peak than does wind. Wind is a close second in cost-effectiveness but is intermittent. Biomass is recognized by DEI as being a baseload generation option and is dispatchable, but is higher cost than wind.
The renewables technologies considered in the resource portfolio optimization model are solar, wind, and bio-methane. Wind is modeled in 50 MW blocks, solar 10 MW blocks, and bio-methane in 2 MW blocks.
DEI believes it is prudent to plan for a Renewable Energy Portfolio Standard (REPS) so each scenario included a REPS. The Reference Scenario assumed a mandated REPS with minimum levels of 1% of total retail energy sales by 2020 and 5% of total sales by 2033. The Low Regulation Scenario has a REPS of 1% of sales by 2020 and 4% by 2033. There is a 1% REPS in 2020 and 15% by 2033 for the Environmental Focus Scenario.
The Traditional Portfolio has 109 MW solar, 35 MW wind, and 12 MW biomass; the Blended Portfolio has 139 MW solar, 178 MW wind, and 14 MW biomass; and the Coal Retires Portfolio has 265 MW solar, 173 MW wind, and 27 MW biomass.
There are a number of issues with DEI’s treatment of renewable energy and DG in the IRP:
- DEI seems to imply that the effect of customer-owned generation is reflected in the load forecast. But it does not indicate how this is modeled, especially when technology is changing rapidly and the costs of renewable energy and DG are falling steadily.
- DEI does not discuss how technological change is causing the cost of DG to fall significantly and how customer attitudes are changing toward the ownership and use of DG facilities. What might the implications be for the utility and how might its resource portfolio change should these circumstances become more pronounced? A thorough discussion and analysis of this topic would have been helpful.
- To the extent the effects of customer-owned generation is not reflected in the load forecast, DEI says the projections for EE and renewable energy can be viewed as placeholders for DG resources. Again, it is not obvious that this is the case given the rapid changes in technology and falling cost for DG.
Indiana Municipal Power Agency (IMPA)
DG and Renewable Energy
Distributed generation was not considered as an option in the resource plan development process beyond a brief general discussion of net metering and other retail customer-owned generation. IMPA knows of six net metering customers and IMPA has a contract with a commercial/industrial customer of one of its members to purchase excess generation from that customer’s onsite generation facilities. The customer has been selling small amounts of energy to IMPA under a negotiated rate. There are no customers that operate a combined heat and power (CHP) system. Based on EPA data, IMPA is aware of 15 industrial boiler installations in IMPA member communities. Nothing is known by IMPA regarding the size or condition of these facilities. With the exception of emergency back-up generators at some hospitals, factories, and water treatment plants, IMPA says it is unaware of other non-renewable retail customer-owned generation in its members’ service territories.
IMPA recognizes that, under the right circumstances, CHP systems would be beneficial to both the customer and IMPA, but notes that the operating conditions and economics must be in place for both parties if a CHP project is to go forward. They also state that most DG systems are small and would have little impact on the long-term. Nevertheless, IMPA declares it will work with their members and the members’ retail customers to investigate the addition of CHP or renewable systems at customer locations (p. 11-128).
IMPA’s discussion of DG is focused on what currently exists and not on how things might be in a few short years given the rapid changes in technology and costs, especially for solar. A more thorough discussion, at a minimum, of the possibilities and implications of greater penetration of DG would have been desirable.
IMPA says it included the following renewable alternatives in the resource expansion modeling:
- Wind – Build (50 MW)
- Wind – PPA (50 MW)
- PV Solar (small facilities at member locations)
- Bio Mass (25 MW)
- Landfill Gas (2.5 MW units in sets of 10 MW)
However, another section of the IRP report says a base case was developed that assumes 21 MW of solar park development over the next seven years. Additional renewable energy additions were left up to the expansion model to determine (p. 6-47).
The ten expansion plans discussed on page 11-132 include a base level of renewables or a high level of renewables. The IRP only says on this page that the two levels were previously discussed in the document. The discussion is not entirely clear about what the base and high levels of renewables are, but there is a strong impression that the renewable energy was hardwired in the optimization model.
Wabash Valley Power Association (WVPA)
DG and Renewable Energy
WVPA discusses in Section II on pages 22-23 how it handles end-consumer distributed generation, with emphasis on the interconnection process. WVPA states that any consumer-owned generator is factored into the IRP either through the inclusion of the resource as a generator or utilizing the generator to offset load as a behind the meter resource.
Landfill gas internal combustion generating units are discussed in Section II on page 10 where it is noted that WVPA has 44 MWs of landfill gas generation capacity and plans to add another 3.2 MW in 2014.
The section of the IRP titled “Selection of Resource Options” does not discuss distributed generation or renewable energy. Nevertheless, the Base Resource Plan shows 32 MW of planned landfill gas generation being added through 2032.
The discussion of DG and renewable energy is minimal and provides no insight as to what WVPA thinks of these resource options, or how technological change and falling costs in this area might impact WVPA’s resource needs going forward. Beyond landfill gas generation, it appears that WVPA gave no thought or consideration to the possibilities associated with the various DG and renewable resource options and how these possibilities might evolve given a range of potential future circumstances.
WVPA did note on page 23 in Section II that the projection of peak demand and energy is adjusted as required to reflect the impact of consumer owned distributed generation, but WVPA fails to explain how it was done.