emix — archive

Definition of a "Price Interval"

From: Girish Ghatikar <[email protected]>
To: [email protected]
Date: Sun, 04 Apr 2010 20:54:00 -0000

Phil, I agree that this would be a good way of moving forward (so are thoughts from Ed Cazalet). On your question of sending a constant stream data for all events. I am not fully certain how the RTP programs will be structured within the service providers. WIthin CAISO there are many RTM prices. Hence, I think we should be independent of how the programs gets structured in future and the data models be flexible enough to allow that. I agree with your other part of the question that the receiving systems (e.g., clients within facilities) should be able to select relevant information. I'd be glad to have a separate discussion on current use of OpenADR within California and its potential within other areas you've discussed. In summary, we've initiated discussion within some of these areas and looking at that opportunity. The PLP was funded by PG&E and generally, our resources/funds are mostly for California programs. I'd be very interested to get other agencies engaged for wider applications. Thanks, Rish Phil Davis wrote: > Girish, > > I believe we are in violent agreement, but I was being unintentionally > obtuse. When I read your example code, my initial impression was that > the price was unique to that DR product and to the time in which it > was called. Based on your subsequent explanation, I believe we are in > agreement that the relevant interval here is time. There are separate > streams of data, "event type" and "price", that are coincident with > the time interval but are not a specific function of it; i.e., all are > independent variables. > > This implies that a control authority could send a constant stream of > data covering all events and times, and that the receiving system > could be configured to select only that data relevant to the local > facility. True? > > Your references to CAISO do inspire another question. OpenADR > together with the studies cited are only California oriented. Our > OASIS members in France (Laurent Guise and Francois Jammes) have > asked if the work being done has any non US centric input. Are you > aware of any international applications? Also, has there been any > affirmative activity to see if OpenADR specs would function within the > market rules of the other US ISO's? ERCOT LaaR, PJM Regulation and > ISO-NE FCM all require automated response and either are in the field > or are close to implementation. It will be difficult to promote > OpenADR as a national standard if it does not encompass the parameters > of these programs. It probably does, but it's politically important > to be able to refer to specific studies. > > Thanks! > > Phil > > ------------------------------------------------------------------------ > *From:* Girish Ghatikar [mailto:[email protected]] > *Sent:* Saturday, April 03, 2010 8:29 PM > *To:* [email protected] > *Subject:* Re: [emix] Definition of a "Price Interval" > > Phil, > > Thanks for bringing this up and I think this is the *issue with > terminologies and NOT the concept of use of time for intervals*. > > Based on my interactions during technology integration of CAISO > real-time market (RTM) with OpenADR, we have seen that intervals could > be used in different contexts. What you have is a definition of "price > intervals" is one element. For example, for CAISO, under "prices" on > -- http://oasis.caiso.com/mrtu-oasis/ -- the intervals are used for > time, price, etc. as seen below by definitions. > > *"Interval Locational Marginal Prices*: Five-minute Locational > Marginal Prices for all PNodes and all APNodes in $/MWh, for each > *five-minute interval RTM*. Posts the LMP, plus the Congestion, Loss > and Energy Components that makes up the LMP." > *"Interval AS Clearing Prices*: Ancillary Services Regional Shadow > Price for all Ancillary Service types for all binding AS Regions and > Sub-Regional Partitions. Posts 15-Minute price relevant to the next 15 > minute *binding interval for RTM* on a fifteen minute basis." > > *While I think the "monetary interval" you have is one important > element for DR*, other element is also the "time interval." For > example, the time interval will determine the end-uses that could be > part of DR strategies and for what duration (or not). The length > (time) and the breadth (kW) are equally important. The price will > determine their willingness to participate and time will determine if > they can and by how much. This was also one of the thing we looked at > the recently concluded participating load pilot (PLP) that we > conducted with CAISO (see below). > > *Open Automated Demand Response Communications in Demand Response for > Wholesale Ancillary Services* > Kiliccote S., M.A. Piette, G. Ghatikar, Lawrence Berkeley National > Laboratory; E. Koch, D. Hennage, Akuacom; J. Hernandez, A. Chiu, O. > Sezgen, Pacific Gas and Electric Company; and J. Goodin, California > Independent Systems Operator. In the Proceedings of the Grid-Interop > Forum 2009, Denver, CO, November 17-19, 2009. LBNL-2945E. November 2009 > http://drrc.lbl.gov/drrc-pubs-auto-dr.html > > Moving forward, I think we should acknowledge this and be clear of > terminologies that are in use and have a certain meaning -- they > should retain their meaning as much as possible. > > Thank you, > Rish > > > > > > Phil Davis wrote: >> Keep in mind that when the ISO's refer to price interval in their >> reports and analysis documents, they are referring to a monetary >> interval rather than a time based interval. In other words, there is >> a $15 price interval between an LMP of $45/mwh and $60/mwh. Most >> usage in analysis is asks questions like "at what price interval do >> large facilities respond to DR signals". These kinds of studies >> influence incentives and subsidies, and may inform rate making in >> regulated constructs. However, in the markets, the actual prices >> themselves are known well in advance of Dr events since they are set >> at auctions. >> >> When modeling/forecasting for grid operations, the ISO's will use >> this research to determine the likelihood of calling an event and the >> cost of that event, given weather predictions at various degrees of >> confidence (of the weather). Because of the usage, I would reverse >> the wording somewhat; i.e., that for a given time interval, there >> should be a price associated with it. Recipients would use this >> data, especially in automated systems either to auto respond or to >> signal managers if prices exceeded a pre-set level. >> >> This happens to be a period of rapid rules change following a >> relative period of stability. Predictions are troublesome at best, >> but there seems to be a national trend toward eliminating specific >> pricing for DR, and paying all parties the generation equivalent of >> the energy product produced. this is not without controversy, so our >> wisest course might be to allow for showing prices for DR and for the >> underlying energy. This would support a value calculation that >> encompasses both the overt value of DR plus the value of the avoided >> purchase of energy for the same time interval. >> >> Phil Davis >> >> ________________________________________________________________________________________________ >> *Phil Davis** | Senior Manager *| Schneider Electric Demand Response >> Resource Center | 3103 Medlock Bridge Road, Ste 100 | Norcross, GA >> 30071 | (: 404.567.6090 | 7: 678.672.2433 | Skype: pddcoo >> *: [email protected] >>