Q&A

with Professor Paul Joskow

Not urgent - but on the context slide I think there is another difference. Old system all technologies had lowest cost if we levelized demand - now this is true for network elements but not for generation - that means competing 'lowest cost' objectives in consumption profiles. Do you agree?

PJ: Sorry, but I don't understand the question. You can email me directly pjoskow@mit.edu.

Given Australia's experience of multiple wind farm tripping following transmission network issues, shouldn't we be looking at some joint distribution of resources and network failures to think about resource adequacy?

PJ: The system operator (or ISO in the U.S.) should be integrating the stochastic properties of wind/solar into its annual analysis off network reliability and then direct network enhancements, including storage, and designate development of the necessary network upgrades. In the U.S. the cost of reliability upgrades is recovered through an ISO transmission tariff pursuant to FERC cost of service rules. Enhancements may be designated to the incumbent transmission owner or through competitive procurement.

Levelized cost of energy (LCOE) analysis are popular b/c they're easy to understand, compare technologies, etc. But they are faulted for not showing the different reliability issues of various technologies. Have you looked at applying ELCC or other considerations to LCOE to improve it as a metric? Thoughts on how you might do so?

PJ: I am not a fan of LCOE for a number of reasons. To make LCOE useful at all for comparing the costs of technologies with different production profiles they must be adjusted to account for the different profiles and the value (price) of energy produced across these different and the value of capacity (LCEE). See my AER paper on LCOE.

What do you think of the potential of vehicle to grid where the storage arbitrage is only one of the value streams?

PJ: This is worth further exploration especially regarding impacts on the distribution system. Since vehicle batteries are potentially dispatchable I can see contracts where they can get capacity payments if they commit to charge/discharge under certain contingencies.

How important do you view the adoption locational marginal pricing for the efficient integration of storage to markets that currently have zonal market designs?

PJ: Very important!

Given the low prices at times of solar generation - is there an economic case for installing solar, without explicit subsidies or revenues from emissions/greens markets?

PJ: It depends on the nature of the subsidies. There is GHG mitigation value even if there is no capacity value. That's where the subsidies should focus. If there were a carbon charge solar would benefit from the higher fossil costs/prices even if no capacity value.

What do you see the role is for major industrial loads (particularly aluminium smelters), not in a daily type of demand response, but as an infrequent shock absorber within the system, and do you have any thoughts for how this should be compensated (i.e. low probability, high impact events)?

PJ: It depends on the industry and how it develops. If an industry can "stockpile" output it's demand may be flexible for longer durations. We are exploring this in connection with the production of green hydrogen by refineries and related industries (with CCS). Since hydrogen can be stored its production can be turned on and off for longer time periods. VOLL can for the basis for payments, but the VOLL would be higher because the duration of the load reduction would be longer.

Why are storage facilities deemed "net generators"? Basic physics says batteries have below 100% efficiency. So they must take from the system more energy than they put in - which means they are net consumers?

PJ: I meant to write "negative generators," though net would be negative. This has come up in the U.S. because some ISOs wanted to treat generators as loads when they charge and generators when they discharge. Under most ISO transmission cost allocation rules this would have led to storage paying more transmission charges. This approach has been rejected by FERC. They are negative generators.

Are there any interactions between the ORDC and ramping services?

PJ: This may now be the case in the development of the ORDC but I am not sure. I think that the best approach is to define ramping as a product to be acquired as are other ancillary services.

Isn’t the coordination between generation and transmission investment decisions more or as important as the design of wholesale markets?

PJ: Yes

Texas - Could the last 5g be done with load shedding or paying demand response?

PJ: With 5g you still get the operational/capacity benefit of having some small amount of dispatchable generation to run at times when net demand is high. If we allow more involuntary load shedding then it would be less costly to the system to get to zero g. Of course it would be more costly to consumers. It's like reducing VOLL in our models.

In the simulations for Texas did you not include any pumped storage?

PJ: We generally assume that the existing pumped storage can continue to operate but that new pumped hydro storage cannot be built. I will have to see if we relaxed that assumption for eiher Texas or the Southeast.

PEM electrolysers can operate in a flexible way according to the variability of RE generation and follow real-time prices. What is the market design for hydrogen industry to actively participate in the market?

PJ: Much less costly electrolysers. Our technical group has another session on hydrogen production and storage using electrolysers this week. The pre-read indicates that it is and will be too expensive. I am a fan of electrolytic since building an electrolyser was my Junior High science projects. It was definitely very expensive hydrogen but that was a long time ago.

Do you have a view on how we get a variety of storage durations into the future system - short term batteries as well as longer term storage? What market designs or incentives might support this range of resources?

PJ: Yes, likely though it depends on the costs and conversion efficiencies. I suspect that the longer duration technologies will require planning and longer term contracts since at least in the U.S. there appears to be only a few situations when they are needed. But when they are need (low wind and solar, high demand) they may be needed to several days.

There is some caution about ORDC for resource adequacy in the NEM because it seems to give generators more revenue but might not lead to more investment in generation. How would you respond to that concern?

PJ: Since the ORDC raises prices it should in theory induce more entry, probably via the price of forward contracts. The high potential ORDC price has also led generators to contract for demand response to hedge forward generation commitments. A generator can lose a lot of money at $9,000/MWh if it has a forward supply commitment, the generator trips off, and it has to make up the difference in the real time market at $9000/MWh. ERCOT has a robust forward contracts market, unlike the other ISOs where the forward markets are illiquid. ERCOT also has a lot of cheap wind, solar (potential), and natural gas (sometimes very cheap). For now wind and solar can fly under the higher prices when gas is marginal. This will change and wind and solar continue to change. Will generators then enter in anticipation of a few very high price hours each year while achieving GHG mitigation targets, whatever the mode of scarcity pricing? I doubt it. We are slowly moving toward hybrid markets as in England and Wales.