IEEE Electrification - December 2022 - 42

bound the capacities curve. Generally, it is sufficient to
linearly interpolate between the minimum and maximum
capacities, as shown in Figure 4. However, for systems
that are nonlinear, other points on the curve can
also be estimated.
On the price side, the mean price is estimated using
historical prices, such as prices from the last 24 h. The
minimum price is the difference between the mean and
the product of k and the standard deviation (v) of the
price, again based on the prices of the last 24 h. The maximum
price is the sum of the mean price and the product
of k and v. Note that this choice of estimating the minimum
and maximum prices is somewhat arbitrary. The
building manager or occupants determine the value of the
scalar " k. " For example, a value of k = 0 maximizes the
demand reduction or energy savings, and, as k increases,
the weighting of comfort increases relative to the weighting
of demand (or energy) reduction. Because the flexibility
of loads is dynamic, these price-capacity curves must
be estimated for each market interval. If the market is
based on a single interval, only one curve is required to
represent the next market interval. For markets that are
based on day-ahead pricing, flexibility must be estimated
for the next 24 h.
The TCC Process in a Building
To illustrate this process, we consider a building with a
VAV AHU. The chilled water and hot water to this system
are provided by a chiller and a boiler, respectively. The
TCC process for this use case includes the following
five different commodities: 1) conditioned air, 2) chilled
water, 3) hot water, 4) local electricity, and 5) local
natural gas. Each commodity requires a separate market.
However, because grid services are only for electricity
and because conditioned air and chilled water are
generated using electricity, these commodities are eventually
converted into equivalent electricity at the building
level and become part of the local (inside the
building) electricity market. Therefore, the local electricity
market constitutes the electricity consumption associated
with conditioning air, the generation of chilled
water, other flexible electricity loads (such as connected
lighting, electric hot water tanks, etc.), and all nontransactive
or inflexible loads (such as plug loads, emergency
lighting, etc.).
Although storage devices (thermal and battery) are
flexible loads, for simplicity, these devices are not considered
in this use case. Because hot water for reheating is
generated using natural gas, hot water use is not converted
into electricity. Therefore, the TCC process includes
only three markets: the " air " market for buying the conditioned
air flow that is necessary to maintain the
required zone temperature; the " Btu " market for buying
chilled water for the AHUs, which is purchased from the
chillers; and the local " electricity " market that represents
all electricity loads that participate in the TCC market as
well as the nontransactive/noncontrollable electric
loads, as illustrated in Figure 4. In Figure 4, the devices
and controllable loads are shown in blue, agents that
represent the DERs are in green, and markets are in the
gray background.
The bidding or market cycle begins with the thermostat
agent, which uses the price-temperature curve that
essentially relates the price to the predetermined
Campus
Node
Campus
VOLTTRON
Market
Distribution
Quantity
Price
Market
Quantity
Price
Transactional
Node
Transactive
Campus
Node
Building
Building
Node
Figure 3. The distributed hierarchical market.
42
IEEE Electrification Magazine / DECEMBER 2022
IEEE Electrification - December 2022

Table of Contents for the Digital Edition of IEEE Electrification - December 2022
