I'm making some progress on the dynamic pricing economics front. This issue impacts our entire approach to smart grid, smart metering, and, ultimately, our ability to make use of plug-in hybrid electric vehicles (PHEVs).
It all boils down to demand response (DR), a utility's ability to adjust consumer demand to align with electricity supply. Effective DR is critical to the success of wind and solar deployments. As such, how we choose to implement DR is at the heart of current smart grid discussion. Here's how I'm thinking about it right now.
The question: What is the minimum amount of data and control that a consumer must offer to its electricity provider at the meter interconnection to empower the utility to confidently understand its demand response load shedding ability with respect to that consumer's immediate energy needs?
The answer: Much of the electricity load in a home can be attributed to a handful of services within it. For the sake of discussion, I'll take what I suspect are the top three: comfort heating/cooling, water heating, and clothes drying. I think I can make each of these loads both responsive to price signals and predictable to the utility company without giving the utility direct control over enabling or disabling the load.
Comfort heating/cooling
A consumer should be able to define his heating/cooling needs in the form of a three dimensional demand curve by time period. If I could set my thermostat based on dollars per day (or similar approach) rather than to a fixed temperature, then specify my hours of occupancy, the thermostat would be able to allocate energy to heating and cooling both appropriately and predictably. A mathematical model showing my heating/cooling demand function could be periodically sent by the thermostat, with consumer consent, to the utility to aid in demand response planning. By aggregating these demand functions in a centralized information system, the utility could construct a single demand response function to tie retail real-time electricity price to heating/cooling load.
Water heating
In my house, water heating load is dominated by the bathroom shower. I assume it is the same in most households. The scheduling and duration of showers (and thus hot water usage) would probably show a relatively predictable demand function with respect to price, if only the clear cost of showering at a given moment were readily available. A computer desktop widget or in-home display could easily convert the momentary price of electricity and water into the price per gallon for hot water, and from there to the price per minute for showering. The display might report that, for instance, a seven minute shower would cost $0.24 right at this moment. By implementing such a solution in limited trials before massive roll-out, the utility could develop a reliable mathematical demand model for correlating price with load shedding ability.
Clothes drying
Following on the idea I laid out for water heating, the consumer could be informed of the momentary price for drying a load of laundry. This could be implemented most easily through a numeric display on the console of clothes dryer appliance itself, but it could also be provided through a computer desktop widget or an in-home display.
By making each of these three energy uses responsive to price signals in a predictable way, a utility company could confidently understand its ability to shed load through control of no more than the real-time retail price of electricity. This control empowers grid operators to make confident use of variable resources like wind and solar generation without depending on fossil-fuel powered spinning reserve capacity or draconian in-home appliance cut-off switches. This replacement "spinning reserve" would be based on well-understood, predictable consumer responses to price signals.
Taking this approach helps to mitigate consumer concerns over privacy protection, and it allows the consumer, not the utility or regulatory agency, to make energy resource allocation decisions within the home.
Moreover, without something like what is described above, the plug-in hybrid car concept cannot be accommodated into the electricity grid. This is a huge point, not be be ignored. (See the footnote below.)
To implement this approach, we need some major changes in the operating reserve rules and disturbance control standards set out by NERC and implemented by the regional grid coordinating councils. We also need the concept of the mathematical model of the consumer demand function to be added to the smart grid standards being developed right now by various groups, including NIST, the OpenSG Users Group, the ZigBee Alliance, and the GridWise Alliance.
--
Footnote
It all boils down to demand response (DR), a utility's ability to adjust consumer demand to align with electricity supply. Effective DR is critical to the success of wind and solar deployments. As such, how we choose to implement DR is at the heart of current smart grid discussion. Here's how I'm thinking about it right now.
The question: What is the minimum amount of data and control that a consumer must offer to its electricity provider at the meter interconnection to empower the utility to confidently understand its demand response load shedding ability with respect to that consumer's immediate energy needs?
The answer: Much of the electricity load in a home can be attributed to a handful of services within it. For the sake of discussion, I'll take what I suspect are the top three: comfort heating/cooling, water heating, and clothes drying. I think I can make each of these loads both responsive to price signals and predictable to the utility company without giving the utility direct control over enabling or disabling the load.
Comfort heating/cooling
A consumer should be able to define his heating/cooling needs in the form of a three dimensional demand curve by time period. If I could set my thermostat based on dollars per day (or similar approach) rather than to a fixed temperature, then specify my hours of occupancy, the thermostat would be able to allocate energy to heating and cooling both appropriately and predictably. A mathematical model showing my heating/cooling demand function could be periodically sent by the thermostat, with consumer consent, to the utility to aid in demand response planning. By aggregating these demand functions in a centralized information system, the utility could construct a single demand response function to tie retail real-time electricity price to heating/cooling load.
Water heating
In my house, water heating load is dominated by the bathroom shower. I assume it is the same in most households. The scheduling and duration of showers (and thus hot water usage) would probably show a relatively predictable demand function with respect to price, if only the clear cost of showering at a given moment were readily available. A computer desktop widget or in-home display could easily convert the momentary price of electricity and water into the price per gallon for hot water, and from there to the price per minute for showering. The display might report that, for instance, a seven minute shower would cost $0.24 right at this moment. By implementing such a solution in limited trials before massive roll-out, the utility could develop a reliable mathematical demand model for correlating price with load shedding ability.
Clothes drying
Following on the idea I laid out for water heating, the consumer could be informed of the momentary price for drying a load of laundry. This could be implemented most easily through a numeric display on the console of clothes dryer appliance itself, but it could also be provided through a computer desktop widget or an in-home display.
By making each of these three energy uses responsive to price signals in a predictable way, a utility company could confidently understand its ability to shed load through control of no more than the real-time retail price of electricity. This control empowers grid operators to make confident use of variable resources like wind and solar generation without depending on fossil-fuel powered spinning reserve capacity or draconian in-home appliance cut-off switches. This replacement "spinning reserve" would be based on well-understood, predictable consumer responses to price signals.
Taking this approach helps to mitigate consumer concerns over privacy protection, and it allows the consumer, not the utility or regulatory agency, to make energy resource allocation decisions within the home.
Moreover, without something like what is described above, the plug-in hybrid car concept cannot be accommodated into the electricity grid. This is a huge point, not be be ignored. (See the footnote below.)
To implement this approach, we need some major changes in the operating reserve rules and disturbance control standards set out by NERC and implemented by the regional grid coordinating councils. We also need the concept of the mathematical model of the consumer demand function to be added to the smart grid standards being developed right now by various groups, including NIST, the OpenSG Users Group, the ZigBee Alliance, and the GridWise Alliance.
--
Footnote
Plug-in Hybrid Electric Vehicles - Another area of concern are PHEVs. This is one technology which completely shatters the central generation to end-consumer power system model that is in operation today. Most aspects of these load and storage devices (in technical terms) are yet defined. Policies and regulations are also not quite complete and certainly business models for energy accounting are sorely lacking. There are several cross-industry groups assessing these gaps and proposing solutions. Utilities are attempting to prepare by obtaining advanced metering devices which are able to perform bi-directional energy accounting, by developing strategies around home area networks and information models, and investigating financial accounting strategies, but much of this work is incomplete.
(Quoted from Smart Grid Standards Assessment and Recommendations for Adoption and Development, Draft 0.83, February 2009, by Erich W. Gunther, et al, of EnerNex Corporation for the California Energy Commission.)
Greg,
Thanks for this post and your site in general. With every article like this, the Smart Grid gets more and more tangible.
Andy
Boston, MA