There is strong motivation to better manage demand-side energy loads and to better follow solar and wind production. Reshaping the hourly loads like air conditioners, water heaters, batteries, and electric vehicles (EVs) will help accommodate more solar and wind generation into the system.
Real time pricing or dynamic pricing is being advocated as one way to coordinate operation of customer devices with the wholesale power system. The prices could be hourly and sub-hourly published day ahead, hour ahead, or intra hourly. The logic of dynamic pricing is simple. When the system is stressed the customer will be sent a high price. When system is in surplus the costs go down and the price to the customer is low. The customers use price signals to adjust thermostats, control water heaters, optimize battery charging and discharging, and EV charging. The goal is to lower cost and improve reliability.
Figure 1 shows a California system load for an April day when the Duck Curve produces morning peaks and a late afternoon ramp up to an evening peak. The figure also shows a price curve that reflects system conditions. The lower panel in Figure 1 shows how a battery, like a Tesla power wall, would respond to the prices. The battery owner with the help of automation would charge when prices are low and discharge when prices are high. The battery operation will decrease the steepness of the ramp because it fills the valley and clips the system peaks. Battery operation makes a profit for the customer and decreases system costs. It will increase system reliability. The coordination of system costs and battery operation is all accomplished with prices. No direct control by the ISO or utility is required. Prices-to-devices seems like a good idea. Why don’t we just speed ahead and implement prices-to-devices as fast as possible? We need load shifting now to accommodate solar and wind. Good as it sounds, we need to be cautious.
One big reason for caution is simple – system stability. Instability is the bug-a-boo of electric system operators. Catastrophic blackouts are the result of instabilities. In order for systems to be stable there has to be appropriate feedback between control signals and the resulting responses. With pure prices-to-devices we get instantaneous load response with batteries and EVs, for example, and slow or no feedback to system operators. This is a recipe for instability.
Figure 2 shows how “prices-to-devices” fits in the system. System operators estimate day ahead prices based on forecasted load. The forecasted prices are broadcast to customers. The “prices-to-devices” are used by the customer (in this case a battery owner) to charge and discharge the battery. The customer can respond to the price signal instantaneously, very fast.The response is instantaneous but the feedback to the system operator is relatively slow or nonexistent. This is a situation that is ripe for system instability. Simulations have shown that in fact a prices-to-devices system will be unstable. See http://www.fortnightly.com/sites/default/files/article_uploads/Markets-3-0-IEEE-Paper-11-7-2011.pdf.
There is a practical solution, transactive energy. With transactive energy we can have dynamic pricing and stability. How this works is shown in Figure 3. In the transactive energy model, tenders replace prices. (A tender implies a binding commitment to buy or sell a limited quantity of energy or transport. A price is not binding. It can be changed with no legal consequence to the seller.)
In the Retail Automated Transactive Energy System (RATES) pilot California customers responded to tenders with transactions. Transactions are a binding commitment to buy or sell. These tenders and transactions provide valuable information to system operators in three ways as illustrated by the feedback loop in the lower part of Figure 3. First, tenders replace prices. The kW quantity or size and frequency of tenders is adjusted by a regulated automated market maker to maintain system stability. Second, real-time transactions provide instantaneous feedback on load changes. And third, forward customer positions provide system operators with valuable forecasting information and stabilize customer bills and supplier revenues. (A position is the sum of customer transactions at any given time.)
In summary, attractive as it is, the “prices-to-devices’ approach to managing load and distributed resources is risky. It could lead to surprise system instabilities. Prices to devices may be a wolf in sheep’s clothing. It looks harmless but could have bad consequences.
A system that uses tenders and transactions can accomplish the goals of prices-to-devices without the risks. To learn more about transactive energy, the RATES pilot, and the path forward see our new book, Transactive Energy in California: A Platform for 100 Percent Clean Energy and Electrification.
f you would like to receive more posts about Transactive Energy and related business opportunities by e-mail, just submit your address here.