The new technical report (Sullivan et al, “Comparative Analysis of Three Proposed Federal Renewable Electricity Standards,” May 2009) from the National Renewable Energy Laboratory reinforces the results of prior reports that the impact of a renewable electricity standard (RES) on electric rates will be quite small. Unfortunately, the report also suggests that the southeast will not produce its own renewable energy, rather its utilities will purchase RECs.
The technical report’s regional perspective is at odds with similar studies by others, and with our own investigation into the potential for renewable energy in the Southeast. I was offered an opportunity to comment on a near-final draft of the report; while some of my critiques were addressed, the fundamental issues with this analysis were not addressed due to an imminent publication deadline.
The three major problems with this report are:
- Biopower is poorly analyzed, or at least poorly documented.
- Transmission is assumed to be built quickly.
- Distributed generation is not adequately considered.
Each of these three factors puts the resource mix that is most appropriate to the Southeast at a disadvantage in the model relative to reality. Unfortunately, there is not adequate information in this report to demonstrate how significant the disadvantage may be.
In summary, the NREL technical report provides some useful conclusions regarding the overall costs of proposed federal legislation to drive renewable energy development. (Bottom line, the costs are so small as to be almost insignificant.) However, the details of the renewable energy development described in this paper are only one possible path forward. The paper does not demonstrate that the Southeast lacks the potential to development renewable energy; it suggests that under a case of rapid transmission development and limited use of distributed generation, the renewable energy market will be dominated by regional wind and solar development.
Shortcomings in the biopower analysis
The NREL technical report (and its predecessor) do not specify any details regarding the source of the resource supply curves for biopower, nor the methods used to incorporate those assumptions into the model. Other than a brief definition of biopower, the “Methodology and Assumptions” section of the report includes no discussion of biomass feedstock methods or assumptions. Digging deeper, the detailed model description indicates that biofuel “constraints regulate the capacity expansion of dedicated biomass and coal-biomass coﬁring plants. Total bio-ﬁred generation is limited by a regional feed-stock supply curve. In coﬁred plants, biomass can contribute up to 15% of the feedstock.” I expected to find the supply curve constraints sourced and described in the model documentation, but there is no section that describes the biopower resource data.
An earlier NREL technical report includes slightly more information, indicating that the supply curves are at the “balancing authority” level and measured in units of MMBtu. Neither report provides the original source for the supply curves, a description of how those supply curves were adapted into the ReEDS model, nor any quality assurance steps that were taken to determine if the results appear realistic.
In contrast, there is substantially more discussion of wind and solar resource model assumptions. The ReEDS model is built on the WinDS model, and NREL is unquestionably the national leader on wind resource analysis. The model documentation referred to above is quite extensive for wind resources. The well-intentioned effort to add solar, geothermal and biopower resources to the highly-developed wind resources leads to mixed results.
With respect to solar, two resources are investigated. NREL created its own Concentrating Solar Power resource database, with a thorough level of model documentation. For solar PV, the NREL technical report provides an explicit discussion of its model approach, “ReEDS did not calculate PV deployment endogenously in these scenarios, but instead used the UCS forecast at the NERC region/subregion level.”
In the earlier NREL technical report, the low rate of biopower deployment is contrasted with the much higher rate of biopower deployment in an EIA analysis. The NREL technical report attributes the different results to higher transmission, siting and intangible costs associated with wind in the EIA analysis. The lower wind cost in the ReEDS model leads to results that emphasize wind rather than biomass. While the differences between the NREL and EIA models with respect to wind and solar resources are discussed in some detail, there is no comparison with respect to biopower resource treatment in the two models.
In my pre-publication review of this report, I pointed out to the authors that the assumptions regarding biopower costs seemed out of line with a number of studies and field experience. In particular, we directed them to the work by University of Florida resource experts which examined the resource costs associated with biopower development in the Southeast. Similarly, the recent Navigant modeling of renewable energy potential in Florida indicated that biopower would cost-effectively meet a high percentage of a state renewable electricity standard. Unfortunately, it appears that the opportunity to comment occurred too late in the publication process to consider these studies.
In summary, the addition of solar and biopower resources to the WinDS model represents the most important set of methods and assumptions on which the NREL technical report’s findings rest. The lack of published documentation and peer review of the model’s solar PV and biopower resource assumptions puts the findings into question.
Transmission – the critical link
The report acknowledges the critical role that assumptions regarding transmission play in the findings, but it does not fully explain the implications. The report assumes that “transmission capacity is built as required . . . Results would differ if construction is delayed.”
Unfortunately, the report does not describe how the results would differ. If transmission capacity is not built “as required,” the feasibility of concentrating solar and wind development in western states would be put into question. I and others urged the NREL analysts to model a comparative case in which transmission development was constrained, to determine where the model might indicate alternative wind development might occur, but this was not attempted.
Without unlimited transmission capability, it seems highly likely that renewable energy would be constructed in the Southeast at quite high levels. First, as indicated in the EIA model results, biopower is built at higher levels when wind resources are considered to be more costly. Second, biopower resources are currently cost-competitive with conventional power generation in the Southeast. Third, utility regulatory policy in Southeastern states is likely to drive a very different result than a model might indicate. Southeastern utilities have highly favorable terms on which to build and operate any type of power plant; regulatory compliance costs such as REC purchases would not be an opportunity to generate earnings and would be less attractive to utilities.
Inadequate treatment of distributed generation and other energy market factors
As the NREL technical report notes, “ReEDS cannot analyze distributed technologies with the same rigor as it does utility installations.” While this comment is directed at the model handling for solar PV, it may also apply to the treatment of biopower in a distributed generation context.
One of the major opportunities for biopower development is small-scale (10-100 MW) biopower plants, often utilizing combined heat and power. These facilities could be located at biofuel refineries, pulp and paper mills, large commercial office parks, or any other site where both thermal energy for heat, cooling, or industrial processes is needed. This technology is commercially available, but due to market barriers in many states, is underutilized from a cost-effectiveness perspective.
Similarly, the report relies on Energy Information Administration projections of natural gas prices. Studies have demonstrated that these projections are systematically biased downward (Consadine and Clemente 2007, pdf format). From a regional perspective, EIA forecasts typically demonstrate very strange behavior, with biopower utilization varying up and down rapidly from year to year. While national energy trends are adequately understood using these models, performance issues hamper the value of using these models to forecast regional trends. For these reasons, the climate study groups in North Carolina, South Carolina and Florida all abandoned use of the EIA energy forecasts and substituted their own state-based forecasts.
Another energy market issue that is not explained in the NREL technical report is the discrepancy between technology cost assumptions (derived from Black & Veatch, unsourced assumptions found in section 5.2, pdf report) with other published energy cost data (for example, Lazard 2009, pdf format). These cost factors introduce substantial uncertainty into the findings.
In summary, energy market opportunities and cost factors that are highly significant in utility regulatory proceedings are overlooked or rely on historically problematic assumptions. The appropriate response to such issues, given the lack of better data, is to conduct sensitivity runs in order to identify the potential range of outcomes that might result from improved modeling.
Sorry, the comment form is closed at this time.