This post is part of the continuing Paths to Transactive Energy series. You can find them all listed by clicking on the matching metatag at the bottom of each post.
One of the most contentious areas of the CTS is how much information market participants and market makers should have about one another. The premises pit functionality vs. scalability, echo the arguments of stateless vs stateful communications, and have ramifications for personal fulfillment and personal privacy.
In any truly transactive market, all transactions are committed. Agents that buy too much power, or power at the wrong time must find separate partners to buy or sell the difference. If your agent purchases for normal power consumption for the week you are on vacation, you pay for the power whether you use it or not. Alternately, a solar producer who commits to power sales on a rainy day must buy power on the spot market to make good on his contracts. The Star Wars character Yoda could have been describing transactive energy when he famously said “Do or Do not; there is no try.’
The minimal market information side is epitomized by TEMIX (Transactive Energy Market Information Exchange). TEMIX uses the most restrictive profile of the CTS. EMIX Agents share no information about capabilities or effects or purposes. Much as a web server can services thousands of clients because it maintains no state information, a TEMIX market can scale to high speed and high volume because of the simplicity of interaction. An idealized TEMIX market is based on peer-to-peer trading.
Many of the most fluid financial markets rely on market makers. A Market Maker may use its own portfolio to complete transactions, or assemble several purchases to enable one sale. Market Makers can maintain knowledge of market participants beyond the capabilities of a single participant. In a deep market, the most significant knowledge is where to find counterparties. Some device capabilities offer services to the Market Maker that are unique to transactive resource markets. These are touched on below.
The purpose of an end node in a transactive resource market is to support the owner or inhabitant of that end node. A commercial facility participates in resource markets to better support the business that is based in that facility. A house participates in a resource market to support the needs and interests of the homeowner. These purposes are private, as are the operation of the systems within the end node.
The systems within an end node can themselves be organized using the CTS, creating a microgrid operated by a micromarket. This market is inherently shallow. There is a limited pool of counterparties at each moment. In a trivial but concrete example, who can sell power to the smart toaster when the occupant pushes down the lever? If a generator must be idle for twenty minutes before being tuned on again, how will the home plan? The specification that defines the CTS also describes how to communicate capabilities, or resource descriptions, using the same semantics as the CTS.
There is no universally correct answer for whether to use resource descriptions or not. Resource descriptions were developed on the models used for bidding into North American bulk power markets. The systems that run these markets are at the limit of their capabilities to handle complexity. They rely on day-ahead markets to allow pre-planning. They will not support a dynamic market with widespread deployment of distributed energy suppliers and purchasers. The deeper the market grows, the more the exchange of resource capabilities are a hindrance to dynamic balance.
A better case can be made for resource descriptions in a shallow market, that of the home and neighborhood, of the commercial building and the office park. For the near future, many systems participating in these microgrids will not natively understand transactive energy. Many systems will still be managed by direct control. Resource capabilities may enable better coordination of a limited number of market participants, i.e. systems that can buy or sell power moment to moment.
The benefits of the simpler “TEMIX” model predominate quickly as systems scale. If there are enough participants in the market, the wisdom of markets produces better results than any central planning. In an unpublished paper, Frank Wolak and Akshaya Jha show that even pure financial participants in day-ahead markets improve overall system efficiency and reduce energy costs by improving forecast quality. Financial participants do not fit into the resource description model, but can integrate seamlessly into the abstract TEMIX model.