Population models for storage types - Pyosch/powertac-server GitHub Wiki

up

The current (April 2016) factored-customer model does not handle storage customers, and does not participate directly in balancing. In other words, it cannot generate RegulationCapacity reports to let the balancing market know how much capacity is available. In order to represent thermal storage capacity (heat pumps, cold-storage warehouses, storage water heaters) batteries, and electric vehicles, some enhancements will be needed. Development and implementation of this feature will be tracked by Issue #863.

Common representation

For all storage types, there is both a minimum load and a maximum load in a given timeslot that must be consumed in order to satisfy device constraints and meet user requirements. This applies thermal storage devices such as water heaters or heat pumps, as well as electric vehicles that individually may or may not be plugged in at any given time. Note that for battery devices that can be discharged into the grid, the minimum value may be less than zero. The key is to project minimum, preferred, and maximum values out to some "planning horizon" that is device-dependent. Regulation actions will change the minimum and maximum values in subsequent timeslots, thereby triggering re-planning. The preferred values may be dependent on prices, in the case where the customer is subscribed to a non-fixed tariff. For example, if the current price is the lowest price anticipated over the device's planning horizon, then the preferred load would equal the maximum load, there would be no down-regulation capacity, and up-regulation capacity would be available only in case the payment for up-regulation was high enough to motivate the customer to shift consumption to a potentially higher-priced timeslot within its planning horizon.

Current structure

Each customer model in the factored-customer world is represented by two structures: a CapacityStructure, and a TariffSubscriberStructure.

CapacityStructure

Contains the configuration elements for some capacity type. This drives consumption behavior, and describes what will happen in the case of a curtailment. It does not contain the information needed to compute regulation capacity, and the shifting behavior is probably not sufficiently detailed to support an electric vehicle population.

TariffSubscriberStructure