Comments

9 responses to “Merchant Power Plant Proposed for Charles City County”

1. 

   tzoser

   October 28, 2016

   In the hearings for Dominion’s 2016 IRP, attorneys from the Southern Environmental Law Center (SELC) were able to get Dominion’s planning director to admit that Dominion’s planning model had a self-imposed limit on how much power it would purchase from others, even if the outside energy was cheaper. This resulted in showing that new Dominion-owned plants might be needed sooner than actually necessary.

   There are good reasons for utilities to own more of their generating capacity. Especially after a long period of relying on a significant share of imported power. One of the big advantages is that owned units can be put in the rate base and earn a rate of return in addition to the recovery of costs.

   The Dominion planning director also told the SCC that the utility hopes to use highly efficient combined cycle units to become a net exporter of power. This would increase Dominion’s revenues and profits assuming the plant could sell enough power at a high enough price.

   According to a Bloomberg article, “the strategy for gas turbine developers in PJM is to jockey for position at the front of the merit order: build the most efficient combined-cycle possible, source the cheapest local gas you can find, undercut your neighbors’ costs (including coal units), and run all day, almost every day, with annual capacity factors as high as 90%. This strategy achieves the lowest-possible levelized costs of gas-fired generation.”

   The Bloomberg analyst goes on to say that things are different in California where the amount of solar contribution is much higher. In 2017, “Bloomberg Fair Value Curves suggest that around-the-clock average ‘spark spreads’ (ie. gas plant operating margins) will land in negative territory. Even the most efficient generators would lose money if they remained online year-round, as around-the-clock average power prices will fail to cover costs of burning gas, bearing a carbon allowance burden, and incurring non-fuel O&M charges.”

   Bloomberg experts were startled by this result. If gas units set power prices virtually year-round in California, how is it possible that power prices could so consistently fall to loss-inducing levels? Wouldn’t generators demand higher prices or simply shut off?

   What they found was “an entirely new paradigm for midday power price discovery in California – one that is intimately related to solar’s diurnal onslaught and to the ‘flexibility’ of the gas fleet. The short version: instead of resting on the short-run cost of gas-fired generation, midday power prices in California are increasingly influenced by gas plants’ operating constraints (like ramping, run-time, and startup costs). In particular, the costs associated with starting up and shutting off a gas unit appear to have the greatest impact on power prices.”

   The new units Dominion is building are 3 x 1 combined cycle units. With 3 combustion turbines (like peakers). The waste heat extracted from the exhaust is used to run one steam turbine (hence 3×1). Incorporating the steam cycle to increase efficiency (55% instead of 35%) makes the overall plant less responsive to ramping up and down and turning on and off – compared to simple combustion turbines.

   Things are different currently in California than in Virginia, but it is an indication of a possible future. Dominion stated in their IRP that they expect natural gas prices to be $6 mcf in 2025, which is more than 3 times the current price. A general rule of thumb is that fuel contributes half of the price of energy for combined cycle plants. To be conservative and to recognize that the new plants will have the highest levels of efficiency, I have assumed that fuel will be just 40% of the cost for the new units. Even so, a 300% increase in fuel costs would yield an energy cost from units such as Brunswick that are 80% higher than today in 2025.

   This is a good deal higher than the current levelized cost of energy for solar, let alone the price of new solar in 2025 which will be at least half of current costs.

   Dominion’s planning model also includes a penalty for grid integration of solar of $390 /kW. This is 40% of the total installed cost of solar expected in 2020. Wherever Value of Solar tariffs have been objectively assessed, solar has provided a net savings to the grid even after grid improvements have been made where required. Including such substantial penalties in the planning model significantly tilts the odds in favor of conventional generation.

   People tend to select the data that favor the outcome they desire. I think Dominion sincerely believes that new combined cycle plants will best serve their customers and shareholders. But if their assumptions prove incorrect, both will suffer.

   From a risk management point of view, it would be better to use energy efficiency or purchased power to meet any increases in demand in the next 10 years while we watch the trends unfold. This will be hard for aggressive utility executives to consider, especially when they are concerned about their revenue stream and share price. They want to be first to the party not last.

   But the assumptions underpinning our national utility system are shifting quickly. With a bit of patience they might find out that the party they were so eager to attend was a bust and they were lucky to miss it. But it will take leadership by the SCC or the Governor to see the bigger picture and protect the interests of both the ratepayers and shareholders.

   Log in to Reply
2. 

   Larrytheg

   October 28, 2016

   excellent analysis – we are pretty fortunate to receive it . thanks!

   but I do have a question somewhat along the lines of your narrative – and that is what keeps these merchant plants from incorporating solar – and basically betting that in using solar when it is “harvestable” but then burning the gas when it’s not – that they can sell electricity cheaper than plants just running gas alone especially if gas gets more expensive?

   question 2 – if more and more of these merchant ventures get built – will they be able to sell power cheaper than Dominions Nukes or are those nukes essentially already paid for and now they’re gravy for Dominion?

   Log in to Reply
3. 

   tzoser

   October 28, 2016

   Larry,

   Question 1:

   That is what Bloomberg was describing in California. The solar does not necessarily have to be at the same location as the gas-fired plant, just somewhere on the grid.

   The contribution from solar gets so large in the middle of the day that it displaces the need for some of the baseload capacity. The baseload plants that can ramp up and down most quickly or the ones that are cheapest to turn on and off are the plants that set the price in the midday market.

   The newest combined cycle plants are a bit more flexible than older combined cycle units and also more efficient so they will set the price. However, with a significant contribution from solar the gas-fired units will not be able to operate as many hours as they would like, so their fixed costs will be distributed over fewer hours which will drive up their overall energy cost. This is what is causing the negative profitability of the gas units in California, even though they are the lowest cost conventional sources of generation.

   If you sited solar and gas-fired units at the same location, both facilities would have to make their own business case. Although, locating a utility-scale solar facility where it could share a substation and transmission lines could lower the cost of that option. However, if objective pricing comes along for the value of distributed solar, that might be cheaper than utility-scale since the central station solar would still have some allocation of the transmission costs (capital and line losses) that are not required for distributed solar nor would the utility-scale solar have the same benefit for grid resiliency as distributed solar.

   Question 2:

   Natural gas-fired generation is giving nuclear plant operators fits throughout the nation. Natural gas-fired baseload units are now setting the clearing prices during much of the day. These prices are well below the prices that were prevalent with a large share of coal plants. This is due to the historically low price of natural gas ( a short-term phenomenon that might not last too long) and the higher fuel efficiency of the new combined cycle units. In deregulated markets such as California, the Midwest (Excelon) and parts of the Northeast, many nuclear units are closing because the lower revenues created by the lower clearing prices from natural gas units are not sufficient to cover the costs of operation. This is occurring even with plants that were purchased at a significant discount from their original owners (such as many purchased by Excelon).

   New York is considering paying a $500 million per year subsidy to keep their existing nuclear units open a while longer. They want to avoid building new natural-gas fired plants as replacements for the nukes. In their view, the natural gas plants will add to CO2 emissions and will soon be undercut in price by renewable sources. New York has a goal of providing 50% of their electricity using renewables by 2030. Building number of new gas-fired units would make that goal nearly impossible to reach. So they are buying time by spending more to keep their nukes running a while longer.

   In California, Pacific Gas & Electric wants to charge ratepayers more to pay for a phaseout of the Diablo Canyon plant ($1.77 billion in added rates). The utility believes that energy efficiency and renewables are the lowest cost replacements for the zero carbon energy from the nuclear plant and they want the funds needed to continue operating the plant until the new sources become available.

   I don’t know the situation with the energy costs for Dominion’s nuclear plants, since they are being bid at $0 in the energy market. Dominion’s nukes might also still be in the rate base and partly funded by the ratepayers, although I don’t know this for certain. One thing that is certain is that the new natural gas generation coming online in PJM will establish a lower energy price and that will reduce the revenues resulting from the operation of Dominion’s nuclear plants. Even though the plants might be mostly or fully depreciated, they still have a considerable cost of operation. Operating expenses, fuel costs, cost to store spent fuel and other nuclear wastes onsite run up quite a bill.

   As usual, California and New York are the first to explore the path ahead. They are both more sensitive to CO2 issues than is the case in Virginia, but purely from an economic point of view they see that the future lies with energy efficiency and renewable generation (mostly distributed). We should learn from their example, rather than rush off and invest in the mistakes both of those states are trying to avoid.

   Log in to Reply
4. 

   Larrytheg

   October 28, 2016

   in terms of gas and solar – and PJM – and companies bidding to provide electricity –

   if a company bids to provide electricity – does it matter to PJM how it is generated?

   In other words if a company owns a site where they have a gas plant and solar – can they commit to provide a certain capacity on demand – and they will decide which “fuel” to use – gas or solar?

   So if that company has sunk costs in the gas plan on that site – they can recover those costs quicker if the electricity they provide costs them less to generate …and they can do that – because they have solar on site and can leverage that solar for higher profits – and buy down their capital investment – for both solar and gas …and once their solar investment is recovered… profits on future solar-generated electricity go to pay off the capital costs of their gas plant.

   So my point here is that a company that bundles solar with gas is going to be more price competitive than a company that runs just gas.

   how have I got this wrong?

   Log in to Reply
5. 

   Acbar

   October 29, 2016

   LG, you ask, “if a company bids to provide electricity – does it matter to PJM how it is generated?” No — with one exception: PJM offers to wholesale customers a “100% renewables” product, and of course energy sold as “renewable” is only sold once.

   Then you ask, “If a company owns a site where they have a gas plant and solar – can they commit to provide a certain capacity on demand – and they will decide which “fuel” to use – gas or solar?” It doesn’t work like that.

   The capacity market is for megawatts committed as much as 3 years in advance which the owner commits can be dispatched whenever the system operator (PJM) so directs. These commitments are sold by the generator owners and bought by the load serving entities, or LSEs (your co-op is an LSE, so is Dominion); each year every LSE must present to PJM a set of commitments for capacity (owned or under long term contract or purchased in the PJM capacity market) equal in total to the LSE’s forecast load plus a reserve amount (currently around 15%, and re-evaluated each year based on the performance reliability and maintenance downtime of all PJM generation). The cost to operate the unit does not matter in the capacity market, but PJM does care which specific units are committed to run when called for. PJM keeps track of the performance of each unit when “dispatched” or called for; if the unit breaks down or refuses to start, that goes into PJM’s history for that unit and eventually the unit could be disqualified from participating in the capacity market if its performance is terrible, but generally “capacity is capacity.”

   Later, in real time, the system operator dispatches all those capacity resources placed at its command based on their day-ahead and short term energy bid price, lowest bid price first. Every generator sold to a PJM LSE in the capacity market MUST submit a bid to run, at some price, or must be declared unavailable for a valid reason (usually because it’s down for maintenance). Generator owners deliver energy and LSEs receive energy as the result of the dispatch process; the PJM energy market is simply the implied purchase and sale of energy that results. The actual energy purchase and sale transaction takes place at the energy clearing price in the market from moment to moment, which in PJM is the locational marginal price. This means a generator owner gets paid, and an LSE pays, the energy clearing price regardless of what the generator owner bid earlier — so, a generator owner could bid below its marginal operating cost in order to ensure that its unit will be selected to run a lot, or above its marginal cost in order to ensure that its unit only runs in extraordinary circumstances, but the best option for most unit owners is to bid at or near its marginal cost. Generator owners also bid their estimated cost to start up and shut down and run at less than full load, all of which goes into PJM’s calculation of which generators to dispatch. From the LSE’s point of view, the fact that it bought capacity from a particular generating unit that did, or did not, get dispatched is irrelevant. From the generator owner’s point of view it’s the primary thing that matters. Most of the larger wholesale market participants wear two hats: they serve end-user load (they are LSEs), and they also have generators they own (they are generator owners) — but PJM treats these market and billing relationships separately.

   So, to answer your follow on question, these units are not “bundled” in any way relevant to a “solar with gas” package for sale. Each unit simply has a value in both the capacity market and the energy market, and that value is not enhanced by “bundling.” Or, to put it another way, everything is bundled as one system through the PJM markets and the least cost options from that entire bundle are selected.

   Log in to Reply
6. 

   tzoser

   October 30, 2016

   Acbar,

   Excellent explanation. Thank you. Yours is a much fuller explanation of the point that I was trying to make by saying each facility (solar and gas) must make its own business case.

   Log in to Reply
7. 

   Larrytheg

   October 30, 2016

   re: bundling

   here’s the scenario

   Company A bids that they can provide one megawatt at a certain time for a certain price.

   As long as they can and do provide that megawatt – does PJM care how they did it?

   Let’s say they have a gas plant that is capable of generating megawatt and the day comes when PJM wants that megawatt “delivered”.

   On that day -the sun is shining out the wazoo and the company says ” we can provide more than 1/2 that Megawatt with co-located solar and save money on gas by only using half as much gas as we would have if we had to generate the whole megawatt from gas.

   so why wouldn’t this “work”?

   this is like “hedging”… right?

   Log in to Reply
8. 

   Acbar

   November 1, 2016

   LG, a delayed response to your question — you ask, “Company A bids that they can provide one megawatt at a certain time for a certain price. As long as they can and do provide that megawatt – does PJM care how they did it?”

   Different answers for different markets: In the 3-year-ahead capacity market, the deals are done in terms of specific units, to avoid double counting, but essentially PJM doesn’t care what unit delivers as long as the seller commits to deliver enough megawatts reliably for the account of the buyer, the LSE. In the day-ahead/real-time energy market, the seller’s bid is not to the LSE but to PJM, the system operator, and it is to operate the specific unit when dispatched for the bid price — that means THAT specific unit with all its specific operating characteristics including startup and ramping times and efficiencies.

   So, “on that day -the sun is shining out the wazoo and the company says ‘we can provide more than 1/2 that Megawatt with co-located solar and save money on gas by only using half as much gas as we would have if we had to generate the whole megawatt from gas.’”. (1) as TZ explained, ‘co-located’ is irrelevant as long as both are connected to the PJM grid, (2) the two components you describe will each submit bids based on their own operating characteristics and costs, and PJM will decide which to dispatch when in order to achieve the lowest grid cost from time to time, and (3) solar operating cost is so low that it’s essentially dispatched whenever it’s available, and of course it can’t be dispatched when it’s not available because the sun don’t shine. Really, the solar and gas ‘components’ you contemplate as a combined unit work just fine totally separately, under one owner or not.

   “This is like hedging, right?” Yes and no. Yes, the generation owner may choose to build and operate both kinds of generation in order to have both income streams; that’s a long-term range financial hedge. No, neither the generation owner nor PJM cares a twit whether these are “packaged” or “co-located” or “more efficient” in sum than individually — because they each will be dispatched when it’s maximally efficient for that specific kind of generation to run at that bid price.

   Log in to Reply
9. 

   Acbar

   November 1, 2016

   Further to LG:. These questions are covering some of the same ground as in the comments on ‘How Fast Is Electricity Usage Growing — a Multibillion Dollar Question’. Won’t repeat all that here, but q.v. — and, nice to have BR up and running like old times once again!

   Log in to Reply