Windpower Discussion on InPowerg

Discussion Archive on ABT and Windpower in inpowerg mailing list: A Preview is given below.

The above archive has been made available with the permission of the list owners and should be a good guide to all.

ABT UI and Wind Power - Discussion in the Inpowerg mailing list

The discussion below, which has been organized based on the time at which these mails were posted to the mailing list is one of the most detailed and honest views expressed by some of the key participants in the ABT and Energy Trading space. This discussion would provide invaluable insight into the issues and options that the availability based tariff regime and the future trading scenario's holds in store with regards to non-conventional generation like Wind, and the many other sources of Non-Conventional generation and dispatching.

Mark Lively on 12/5/2007 12:10 AM IST

I don't know that you can "solve" ABT, or the UI pricing portion of ABT, which I mostly talk about. UI pricing ends up being a control mechanism, part of the negative feedback that gets the system back to where it is desired to be.

I think that the UI pricing needs a few tweaks or modifications. First, there should not be a ceiling. As frequency declines, the price should continue to increase. The importance of this is that it allows an investor to build a power plant and earn sufficient revenue during times of low frequency to pay for the investors capital cost. If there are enough periods of low revenue and the effective average price is high enough then more investors will build power plants. With more power plants the frequency will go up.

Second, the curve should be automatically changed based on time error. As synchronous clocks lag behind asynchronous clocks, the pricing curve should be increased. If the synchronous clocks are ahead of the synchronous clocks, the price curve should be lowered.

Third, the price for reactive power should be based on both the price for active power and on local voltage conditions. ABT has some voltage component in setting a payment for reactive power, but it needs to be refined.

The ABT should also have a geographic differentiation component. Some of the geographic differentiation will be part of the reactive power price.

I attach

• “Reply Comments Of Mark B. Lively In Regard To Substituting Formulary Markets In Place Of Cost-Based Rate Schedules For Reactive Power,” Market-Based Rates for Wholesale Sales of Electric Energy, Capacity and Ancillary Services by Public Utilities, FERC Docket RM04-7-000, 2006 September 20.
• “Better Pricing Of The Energy Balancing Market,” 27th USAEE/IAEE North American Conference, Houston, Texas, 2007 September 16-19.

Mark Lively on 12/4/2007 9:08 PM IST

I am not sure how people are defining the balancing market. I would tend to say that the UI (Unscheduled Interchange) market in India is a balancing market. But, based on the discussion, some people also say that a balancing market involves pre-arranged deals between two (or more) parties to meet their anticipated imbalances. The difference between these anticipated imbalances (which are now scheduled, and prearranged, and not quite imbalances any longer in my mind set) and the actual imbalances would be the UI market. This latter confusion is why I don't like to think of pre-arranged deals as including something called a balancing market. It gets too confusing. I want to just talk about the UI market.

Given that this discussion thread is about a UI market within Maharashtra, I am going to introduce some confusion by talking about two UI markets, the Maharashtra UI market (MUI) and the CERC regulated UI market (CUI) between Maharashtra and the rest of the grid. CERC has set the pricing plan for CUI. I have stated at various times and venues that the CERC pricing plan can be improved by making the pricing curve steeper and responsive to longer term market conditions. There also needs to be geographic differentiation and an improved payment for reactive power.

In reading the MERC guidelines for MUI, I disagree with the utilities that the price paid should vary directly with frequency. The correct driver is Area Control Error, which is Maharashtra's CUI with the rest of the grid biased by the frequency error. Yes, the MUI price should go down as the frequency goes up, but the MUI price should also go down as Maharashtra's CUI becomes increasingly positive. A high frequency and a positive CUI are both indicators of a surplus of electricity. Set the price for MUI using ACE, not just frequency.

This thread of comments was started by Prasanth Gopalakrishnan when he commented

The MERC guidelines for Intrastate ABT is based on System Marginal
Pricing (SMP) rather than on frequency based UI charges?.

In some respects, the existing CUI pricing can be considered to be SMP. Certainly CUI pricing is for the system and prices only marginal deliveries, since most non-marginal deliveries are priced in another manner. The MERC order discusses Ex Ante SMP, prices that the Maharashtra State Load Dispatch Center (MSLDC) produces ahead of time, as well as Ex Post SMP, prices that the MSLDC produces after the fact, at least as I read the order of 2007 May 17. I presume that the MSLDC produces a dispatch schedule associated with the Ex Ante SMP, such that participating parties have an obligation to delivery specific quantities at specific prices.

Most Ex Ante SMP programs assume that they match generation to load exactly, with zero frequency error and zero UI. Obviously there is frequency error when the system is actually run and there is UI when the system is part of a larger interconnection. With these insights, the Ex Ante SMP can be viewed like the CUI price when the frequency is 50 Hertz.

Most such Ex Ante SMP are internally consistent, such that the price at one generator relates to the price at another generator by marginal line losses and line constraints. With such an internally consistent Ex Ante SMP, a good way to produce Ex Post SMP for pricing MUI is to adjust the Ex Ante SMP by a formula based on Maharashtra ACE. The formula would be like the CUI pricing curve except that the price at 50 Hertz (actually at zero ACE) would be the Ex Ante SMP instead of being fixed. The time varying null price at 50 Hertz would be an improvement on the CUI pricing mechanism. A detriment to this time varying null price is that it is time varying, destroying some of the simplicity of CUI pricing.

The formula (or pricing curve in the language of the CUI) would encourage generators to move in ways that the MSLDC desires them to move, not necessarily in the way that the Southern Region LDC might desire. Thus, the Maharashtra UI with the rest of the grid may produce pricing forces that are inapposite to the pricing forces of frequency error.

Using a formula based on ACE to adjust the Ex Ante SMP can provide the market participants more confidence that the price they pay or receive for MUI is fair. CERC's creation of the pricing curve for CUI reduced the squabbling about which utility was leaning on the system and which utilities were being taken advantage of. If a party thought the price was too low, it could reduce production, reduce its internal marginal cost toward that too low price. This action of reducing production would reduce frequency and raise the price for UI. Using ACE to drive a formula converting Ex Ante SMP into an Ex Post SMP can similarly reduce the squabbling within Maharashtra.

I discuss an appropriate formula in the attached “Creating an Intra-Hour Market on Bonneville Power Authority As Part of the Wind Integration Rate Case”, 2007 October 9, BPA-Transmission Services | 2009 Wind Integration Rate Case, http://www.transmission.bpa.gov/Business/Rates_and_Tariff/2009WindIntegRateCase.cfm. Equation (1) on page 1 is

Price = Seed * Base ^(-ACE)

In the context of the Maharashtra situation, the Seed would be the Ex Ante SMP, or the price without any adjustment. When ACE is exactly equal to zero, the use of ACE as an exponent reduces the price in Equation (1) to the Seed, neither greater nor less.

I look forward to hearing reaction to the above.

Mark B. Lively

Soorya Shrestha, 12/4/2007 6:06 PM IST

If the balacing market, which is operated by the system operator, is mixed with UI mechanism, which is governed by the market only, then there is possibility of market power being exercised by the system operator, who could turn out to be not independent. The reason is he got the authority to manipulate the market demand by supplying or withdrawing part of the balancing power, which will ultimately change the frequency. If the opportunity of changing frequency is given to any market player, then that will be the source of market power. Then the whole beauty of UI will be lost.

I think this will give opportunity to the system operator to possibility of gaming. So I think they should not be mixed.

Soorya

Raja Rao 12/4/2007 5:11 PM IST

Hi,

When there is an imbalance the frequency is expected to change from the normal of 50 Hz. What is the quantum of deviation of frequency dependant on and is it possible for the system operator to influence this quantum?

-rajarao

Prabuddha Banerjee 12/4/2007 3:45 PM IST

There are two different schools of thoughts on balancing mechanism. The first school or it is probably most widely used thought is that the frequency must be fixed and to maintain that fixed frequency the system operator will sale or purchase power in real-time to meet the balancing requirement. In such case you may find a difference in the volume of unscheduled off-take and unscheduled supply , as balancing power is being purchased in a separate market. This system is in operation in NETA, Nordpool and in many other countries. Here the cost of purchase of balancing power is raised from the grid participants in the form of system sale /buy price, where buying and selling price in most of the time has a difference. System buy or sale price is determined from the market clearing price but with an off set of the money spent for buying/selling the balancing power. So, the offset is more when the system is in highly shortage condition, the offset is low while there is low or no shortage.

The second school of thought says that the system operator does not buy or sale balancing power; the pool participants actually complement each others demand with a unique price signal which is frequency. In this case you will never find any difference between unscheduled off-take and unscheduled supply (barring the loss), because there is no separate mechanism to buy or sale balancing power. This is what is known ABT mechanism in India. ABT determines the system marginal price based on the frequency and the frequency is being determined by the system demand, because frequency may be allowed to vary within a margin.

It is very difficult to say that how the thing will move if both the school of thoughts start to coexist in the same market. In the first system the participant will have hardly any signal to take part in a balancing mechanism, the balancing is the sole responsibility of the system operator, and there is a structure of how the system operator can run such show. Here in the ABT mechanism the operator cannot really authorized to buy or sale any power. Even if you give that power to the system operator you need a real-time market place to buy the balancing power, how to create that, that will definitely take time. I am a little bit confused, how the market is going to be like in this multidimensional market where there are some power exchanges, there are bilateral contracts and normative contracts, multiple forms of balancing mechanism, on the other hand energy suppliers are distribution companies with open access freedom for the consumers( I feel it stupidity on the part of a distribution company if it ever let anybody manage an open access ), transmission system operator-system operator-power grid etc. Who knows what is next..

Prabuddha Banerjee

Sreekumar N on 12/3/2007 11:27 PM IST

Hello

A narrative commentary on the Andhra Pradesh Power Sector in the Nov 17 issue of the Economic and Political Weekly.

http://epw.org.in/uploads/articles/11229.pdf

Looking forward your comments.

Regards

Sreekumar N
Prayas Energy Group

Prasanth Gopalakrishnan on 12/3/2007 9:09 PM IST

http://mercindia.org.in/orders_2007.htm

In SMP, they calculate the cost of additional power drawn based on a Merit Order Rating (MOR) of the most costly power first.

There are possibilities where some of the participants depending on the scenario to pay a price for some-one else's problems.

Sincerely,

Prasanth Gopalakrishnan

Mark Lively on 12/3/2007 2:55 AM IST

UI pricing is a form of SMP. So there may not be a difference. To the extent that there is a difference between UI pricing and the SMP that comes from the MERC guidelines for Intrastate ABT, then the utility that deals with both the interstate ABT of CERC and the intrastate ABT of MERC will experience an automatic arbitrage, buying at one price and selling at the other price. Note that these prices might be structured such that the utility will loose money or such that the utility will make money.

One of the issues should be recovery of the real time electrical losses incurred by the utility between the two points of delivery/receipt. I say real time since sometimes the utility will be experiencing paper losses that are converse to the actual losses. For instance, consider the situation that the actual flow is from the central transmission lines to a load pocket within the utility. But a generator in the load pocket is selling electricity to the utility, and the utility is selling UI to the central system, because of loop flow through the utility to another utility. The price in the load pocket should be higher than the UI price but under some SMP programs the direction of the nominal flow from the generator to the central system would suggest that the price at the load pocket would be lower than the UI price.

What is the URL for the MERC order on SMP?

Mark Lively

Prasanth Gopalakrishnan on 12/2/2007 2:32 PM IST

Hi All,

The MERC guidelines for Intrastate ABT is based on System Marginal Pricing (SMP) rather than on frequency based UI charges?.

Any comments on issues related to practical implementation, and the overall impact in achieving the CERC objectives?.

Sincerely,

Mark Lively on 11/27/2007 10:38 PM IST

I mentioned UI pricing in the article "Microgrids and financial affairs - creating a value-based real-time price for electricity" which I wrote this summer for publication in the Cogeneration and Small Power Production magazine, Vol 8, Issue 5, September 2007. You can download a copy from

http://www.cospp.com/articles/article_display.cfm?ARTICLE_ID=307889

S K Soonee on 11/3/2007 10:42 PM IST

The 4th Annual meeting of Very Large Power Grid Operator (VLPGO) CEO meet held on 29 & 30th Oct 2007 at New Delhi was hosted by POWERGRID India.

Brazil, Korea, Russia, USA, China, Italy and Japan were the countries represented in this meet. The utilities from Spain, France, UK, California ISO USA and MISO USA are also part of the VLPGO.

In an international recognition for India, CMD POWERGRID was nominated the first President of VLPGO for the year 2007-08.

The VLPGO initiative aims to address the concerns of large power grids such as the reliability and economy of operations, efficient market operations, and interaction with other stakeholders besides the technological and human aspect of power system operation. It aims to exchange the technical expertise and best practices adopted by very large grid operators and build a roadmap for the next practices.

The VLPGO has recognized that the electricity grid of the 21st century would need to have a radical change in approach considering depletion of resources, environmental concerns, need for energy conservation and the increasing role of new and renewable energy resources in the grid. The Control Centre Energy Management System (EMS) architecture would also need to be open and accommodate the requirements of the competitive electricity market rather than being solely vendor driven.

Three Working Groups (WGs) on Synchrophasor technology, Prevention of cascading blackouts through a self-healing mechanism and Incentive mechanisms for new capacity addition was constituted at this meet. Workshops on SMART grid philosophy would be held later this year.

The next annual session of the VLPGO would be held in Oct 2008 at Brazil.

Prabuddha Banerjee on Mon Oct 8 20:24:31 IST 2007

When most of the energy that we consume is from fossil fuel, a little improvement in that technology would really change the earth a lot, Other than CCGT technology (H technology - probably the most efficient till date) I feel there are a lot can be done in the basic area. Most of the heat spills through stack; which cannot be captured due to human knowledge deficiency. The stack temperature needs to be kept above certain value depending on sulphur content of the fuel(mainly coal).There are some technology which demand that it can reduce the stack temperature very close to atmosphere, this is called condensing heat exchanger or CHX. If that can be done in a Thermal power plant the efficiency can be improved to a great extent. I really don’t know whether anybody has implemented such technology in reality. If that can be done it can really change the world, because coal probably the most common fuel used in power plant and has a high Sulphur content.

Mark Lively on Mon Oct 8 06:32:13 IST 2007

This is an interesting analysis.

I saw another analysis that described the trade off between wind generators, single cycle gas turbines, and combined cycle gas turbines. The premise was that the output of combined cycle gas turbines would be replaced by a combination of wind generators and single cycle gas turbines, the latter because they could ramp up fast enough when wind died, whereas combined cycle gas turbines could not. The relative efficiencies of combined cycle gas turbines and single cycle gas turbines were such that when combined with the relative hours of operation that the total fuel consumption was the same, that is, ramping up and down single cycle gas turbines to meet the wind outages used as much fuel as if combined cycle gas turbines were operated around the clock.

From what I saw of the numbers, the approximations came out very close, perhaps too close not to have been planned. But the heat rates looked plausible for the turbines and the capacity factor of the wind turbines were in line with what I had heard.

Sebastian Morris on Sun Oct 7 10:23:29 IST 2007

The green aspect of wind is overstated. If the subsidy given to wind, is instead given to thermal to step up efficiencies by even 6% the impact on carbon reduction is much greater and more sustainable. Thus the US initiative on the Advanced Turbine Project does more to reduce carbon emissions.

Soorya Shrestha on Fri Oct 5 18:22:17 IST 2007

I would say the Wind energy is getting correct price in UI mechanism. That is the beauty of UI. Wind energy is a new participant in the ABT and even in power system itself. The approximate similar plant compared to wind energy is the run-off-the-river hydro power plants. They also generate when the water is available. Like wind in atmosphere, the discharge in the river controls their output.

The water discharge is not in control of the operator. But the forecast of river discharge is very precise. So wind power participants in power system also should make correct forecast. That is the part of the developers' job, not the system operator or the regulators job. As done by hydro plants, yes you are right they should maximize their benefits with carbon credit or other innovative ideas. For example, they can make provision for tourists to visit the top of the wind power generator with lifts, restaurants etc. I think no wind power plants have considered this idea. There could be tremendous revenue from tourist attraction. There could be so many other benefits with this mechanism.

In short, promoting wind power plants is the job of promoters. No additional benefits need to be given to any type of plants including wind power plants. Whatever UI provides is the actual price of the energy in real time. Suppose, the wind power plants gain benefits with tourist attraction, they will not share that benefit with the power system! Regarding fossil fuel plants, environmental regulations will bring them to competitive price.

In the end every type of plants has got some risk of getting projected benefits. After all it is business.

Prabuddha Banerjee on Fri Oct 5 14:49:40 IST 2007

I wonder why such diminutive volume of (compared to the total Indian grid capacity) wind energy can have any big effect on the grid. I don’t think the % share of wind energy may change to a great extent in near future. If a good no.s of unit run on Free Gov. that is enough to pacify the effect of unpredictable wind energy. And in ABT regime that would be done in a very economical way. But to me the problem is elsewhere, do the wind firm get the correct price in the process (ABT)? Because wind does not know when the UI is high (whatever UI curve you impose on) so very often the wind firm may deliver energy at low cost and some other generator will make the profit out of it (by reducing generation). In my view a carbon levy needs to be imposed on the consumer to make such green energy economically viable. Because the only advantage the wind firm has in its side is carbon, it should get that price. Or else we have to wait for the fossil fuels to end.and then we will have only green energy and Negawatts.

Bhanu Bhushan on Wed Oct 3 14:45:55 IST 2007

With reference to Mark Lively's comments of October 3, I feel that Raghuram's apprehensions are misplaced.

As a load dispatcher, he should be happier when he is not required to schedule wind generation, which is so unpredictable.

Europeans have a major problem because tie line flows have to be maintained as per schedules. We have no such problem because we permit deviations.

A Preview is given above: For full discussion Archive on ABT and Windpower in inpowerg mailing list, please follow the link: