In an earlier post I addressed the potential availability of power to replace Indian Point’s capacity. This is an update to that analysis with a discussion of a new report. In January 2017 New York’s Governor Andrew Cuomo announced the closure of the Indian Point Energy Center located 25 miles north of New York City. Cuomo claims that Indian Point produces 2,000 megawatts of electrical power and that “more than enough replacement power to replace this capacity will be available by 2021”. Since then environmental advocates have claimed that the replacement power “can and must be replaced with a portfolio of energy efficiency and clean energy resources: renewable resources such as wind, hydroelectric and solar.”
The basis of this claim is the Synapse Energy Economics report prepared for the Riverkeeper and Natural Resources Defense Council entitled “Replacement Energy and Capacity Resources for the Indian Point Energy Center Under New York Clean Energy Standard (CES)” that claims that replacing Indian Point can be done with a combination of renewables and energy efficiency. This post addresses that report.
I am generally skeptical of the Synapse report for two fundamental reasons. The report is sponsored by organizations that don’t want Indian Point or central power stations in general and also support development of renewable energy. It would not be presented to the public if it did not support their goals and objectives. Secondly, the analysis uses the National Renewable Energy Laboratory (NREL) ReEDS (Regional Energy Deployment System) modeling system. Clearly this organization has specific goals in mind and I doubt that they would have developed a model that did not support their renewable energy goals. These reasons do not necessarily say that the results should be dismissed out of hand but it does suggest that the results have to show no signs of bias towards a pre-determined result to be valid.
In order to determine whether there is bias it is necessary to dig into the approach. ReEDS is a” long‐term capacity expansion and dispatch model of the electric power system in the lower 48 states”. The report notes that it has a “high level of renewable energy resource detail with many wind and solar resource regions, each with availability by resource class and unique grid connection costs”. Capacity, dispatch and load planning models are complicated and the characterization of the grid is a particular problem for New York. EPA and RGGI have relied on a similar model called the Integrated Planning Model that is notorious for its mis-characterization of the New York grid with the particular problem being transmission constraints to New York City and Long Island. EPA uses a version that has yet to get it right but New York’s influence in RGGI has led to a version that is acceptable in this regard. In the absence of a thorough review of the model I doubt that ReEDShas been tuned to correctly characterize this problem. If not, the fact that there are transmission constraints can lead to poor estimates of future load and capacity development making it easier to claim that renewables can be integrated easily.
The bigger problem is that the modeling approach assumes too much. In particular, they assumed that the levels of energy efficiency necessary to meet the CES will be met. The report itself describes the weakness of that assumption: “The CES order assumes annual incremental savings through energy efficiency of roughly 1.5 percent of overall electric energy demand, resulting in a reduction from ~160k GWh in 2016 to ~146k GWh in 2030. However, the CES order does not include any mechanism to ensure that these levels of energy efficiency are achieved, in contrast to the binding and enforceable 50 percent by 2030 renewable energy target enacted by the CES order. Nor has the Public Service Commission enacted any other policies outside the scope of the CES order to ensure that the state achieves these levels of energy efficiency. Rather, existing policies (which consist primarily of Energy Efficiency Transition Implementation Plan (ETIP) targets and budgets for each of the state’s investor‐owned utilities) guarantee only a small fraction of this 1.5 percent annual incremental savings.”
In other words, the modeling assumes that energy efficiency gains will occur despite the lack of a mechanism for it to occur. Furthermore, the insinuation that the renewable goal will be met because there is a binding order misses the point that the goal may not be achievable.
I am a fan of energy efficiency because it has no regrets. In other words because there is no completely benign way to make electricity reducing the amount you need is a very good thing. Moreover, you can target the energy efficiency and conservation funding so that those least able to pay for their energy get direct benefits. As a result even if it turns out that we don’t need to reduce CO2 emissions because of its impact on climate we still get benefits from this approach. Unfortunately I have qualms about the capability of energy efficiency to provide a significant amount of replacement power for Indian Point.
First and foremost is simply accounting for energy efficiency. The Clean Energy Ministerial discusses the obstacles to assessing energy efficiency gains and notes “Because savings represent the absence of energy use, it is impossible to directly measure energy efficiency impacts”. In the absence of direct measurements it is therefore necessary to estimate energy use in the absence of the energy efficiency savings project. I would hope that the CES order clearly defines the metrics for these estimates so that they are credible, certain and consistent. Frankly, the lack of documentation thus far in the CES does not give me hope in this regard.
One aspect to the continuing aggressive efficiency goals that I have not seen addressed is market saturation. A significant fraction of New York’s RGGI investments have been allocated to energy efficiency and those investments were preceded by years of subsidies for energy efficiency in other programs. An anecdote is just to consider household conservation. Once all the windows and doors have been replaced by more efficient versions, investing in even better ones will not be as cost effective. After you have spent money to do the obvious things any future investments cost more but yield less relative improvement. My point is that the Synapse projections of further efficiency goals do not include assessments of what is available. Instead they simply assume that the CES goals will be met and that even more reductions are possible if we throw even more money into the programs.
Finally there are concerns about the “rebound effect” that suggest that maintaining the proposed level of annual incremental savings is ambitious. The “rebound effect” is when an improvement in energy efficiency triggers an increase in demand for energy. The impact of this effect is controversial and there is much uncertainty regarding the magnitude of rebound effects associated with energy efficiency improvements.
The renewable goals are similarly based on assuming that the CES requirements will be met. The Synapse report states: “We incorporated New York State CES parameters into our projection of load and resource requirements. These parameters include CES‐assumed increased levels of energy efficiency and meeting the 50 by ’30 renewable energy requirement. We reflect New York’s target of obtaining 2,400 MW of offshore wind energy by 2030 in all scenarios, staged to reflect 600/1200/1800/2400 MW attained by, respectively, 2024/2026/2028/2030.” In 2015 3,398 MW of new offshore capacity was added worldwide, bringing the total to over 12,107 MW according to the Global Wind Energy Council. In theory that should mean that New York can install 2,400 by 2030 but it should be noted that New York’s total on-shore wind capacity in 2015 was 1,891 MW.
The biggest failing in this report is the lack of cost data. The report notes that their cost comparison is “not meant to be definitive in the absolute sense; rather we use a consistent framework across the different scenarios in order to ascertain relative cost patterns”. This modeling purports to show that their preferred alternative is cheaper than other choices using their relative costs. What is missing is the overall cost. The fact that the ReEDS model which “builds” off-shore wind capacity based in part on economics had to be adjusted by Synapse to “hard-wire” the capacity built to match the CES requirement suggests that ReEDS is not as optimistic about that resource. Furthermore, it has been noted that “The average household in Germany contributes an estimated 240 euros a year to renewable energy subsidies.” is not reassuring in this regard.
I will conclude this post with some particular issues with the modeling that may indicate problems with the results. The model “incorporates the addition of previously committed gas‐fired generation in the region in 2018” and cites the “CPV Valley unit, at 650 MW; and a generic combustion turbine unit at 90 MW”. The CPV unit is expected to be online in 2018 but instead of 90 MW of new combustion turbine capacity the Cricket Valley station alone is 1,100 MW This plant is permitted, under construction and expected to be on line in the first quarter of 2020. That means that renewables will not be displacing old inefficient fossil-fired generation but the latest and most efficient fossil technology. Another assumption used in all the modeling scenarios is that there will be a 2.5% decline in the RGGI cap which has been proposed by Governor Cuomo but is not yet RGGI policy.
When I model something the first thing I check is the results relative to recent observations. I have a couple of problems with projections for 2016 compared to actual values. For example, the model’s estimate of 2016 CO2 emissions is 20% higher than actual emissions. When a model over estimates a key parameter by 20% in the first year I don’t put much weight into their projections in 2030. Another issue is the 2016 wind generation estimate of 5 TWh when the 2015 observed wind generation was only 4 TWh. It is highly unlikely that the wind generation will increase 20% in one year so that is another likely over-estimate.
I have posted three tables with model projections for all six model runs for 2018, 2022 and 2030. Synapse Synapse Energy Generation and Capacity Appendix A Scenarios describes the model runs. Synapse Energy Generation and Capacity Appendix A Comparison of Generation Estimates describes the generation (TWh) estimates and Synapse Energy Generation and Capacity Appendix A Comparison of Capacity (MW) Estimates describes the (MW) estimates for fourteen different source-type categories. I have issues with some of those category results. The scenarios address the Indian Point retirements so the nuclear estimates show that. Note that those projections assume no changes in the upstate nuclear units. Cuomo’s war on coal is reflected in shutting down the remaining coal plants. One of the unintended consequences of the renewable and energy efficiency is the effect on the profitability of the remaining fossil stations. For example, in the gas category there is a projected reduction in gas capacity of 21% from 2018 to 2030 in Scenario A6 but the generation drops 74%. My concern is that the drop in the generation from those facilities will mean that they cannot remain viable without a capacity payment. For new gas note that the exclusion of Cricket Valley means that their 2022 estimate of just over 5 Twh of generation is less than half of what I expect if Cricket Valley runs as expected. The wind estimates are all at least three times the observed 2015 NYS generation which confirms my expectation that aggressive wind development is necessary in these projections. It is also not clear why the wind generation projections in all four scenarios are less than either reference case. Even more vexing is that solar in the last scenario does not increase from 2018 to 2030. I am guessing that the solar money goes to aggressive energy efficiency. Note that the modeling assumes several fold increases in DG PV capacity. The last category that I want to address is oil-gas-steam. New York is unique in its reliance on this source category as backup for emergencies. Unless a production cost model is specifically tuned to New York then this is an easy category to turn down. However, reducing the generation, much less the capacity, is not as simple as it appears on first glance because of the role these units cover. I do not think those category estimates are realistic.
Ultimately this modeling exercise is a good example of Pragmatic Environmentalist Principle 4: We can do almost anything we want, but we can’t do everything. In the absence of absolute estimates I can only guess what the ultimate costs will be but the German experience of 20 euros per month for renewables is not comforting. I cannot endorse this approach because I fear the additional costs of renewables will divert too much of the state’s resources relative to other needs. New York State has to invest $40 billion in its water infrastructure just to provide clean water and treat wastewater. In my opinion investing in that immediate need and energy efficiency is a more appropriate social policy than subsidizing renewable energy. If, in fact, Indian Point has to close I suggest that accepting that its replacement power will have to include fossil generation is necessary.