Glossary

← Learning Hub Reference Document · Battery Optimizer Course
Term Definition
aFRR Automatic Frequency Restoration Reserve. Activated automatically by TSO signal to restore frequency to 50 Hz. Full response within 5 minutes.
Arbitrage Buying energy when cheap and selling when expensive. For batteries: charge low, discharge high.
Big-M A modeling trick using a large constant M to link binary variables to continuous constraints. Used to prevent simultaneous charge/discharge.
BMS Battery Management System. Hardware/software that monitors cell voltages, temperature, SoC, and enforces safety limits.
Capture ratio Realized revenue / perfect-foresight revenue. Measures how much of theoretical maximum the system achieves. Good: >0.7.
Constraint A rule the optimizer must respect. Physical limits (power, energy) or contractual obligations (delivery commitments).
Day-ahead (DA) Auction-based electricity market where energy for the next day is traded. Gate closure typically 12:00 D-1.
Decision variable A quantity chosen by the optimizer (e.g., charge power per hour). Contrast with parameter (given input).
Degradation cost €/MWh penalty representing battery wear per unit of throughput. Derived from replacement cost / lifetime cycles.
Delta (Δ) The difference between new optimal schedule and existing positions. The trades needed to rebalance.
EPEX SPOT European Power Exchange. Operates day-ahead and intraday markets for DE, FR, AT, NL, BE, and others.
FCR Frequency Containment Reserve. Fastest balancing product — full activation within 30 seconds. Capacity-based payment.
Gate closure Deadline after which bids can no longer be submitted or modified.
HiGHS Open-source LP/MIP solver. Fast, free, no license. Suitable for battery optimization at production scale.
Intraday (ID) Continuous electricity market for trading close to delivery. Pay-as-bid. Gate closure 5–60 min before delivery.
LP Linear Program. Optimization with linear objective and constraints. All variables continuous. Solvable in polynomial time.
MCP Market Clearing Price. The uniform price at which supply equals demand in an auction.
mFRR Manual Frequency Restoration Reserve. Manually activated by TSO. Full response within 12.5–15 minutes.
MIP Mixed-Integer Program. LP with some integer/binary variables. Harder to solve (NP-hard) but handles discrete decisions.
Objective function The single number the optimizer maximizes or minimizes. For batteries: typically revenue minus degradation cost.
Parameter A given input to the optimization (price, capacity, efficiency). Not chosen by the optimizer.
Pay-as-bid Pricing mechanism where each participant pays/receives their bid price. Used in intraday continuous.
Price-taker Assumption that your trades don't affect market price. Valid for small assets. Simplifies optimization.
Pyomo Python optimization modeling library. Translates mathematical models into solver-readable format.
Robust optimization Maximizes worst-case performance across scenarios. Conservative but protected against bad outcomes.
Rolling horizon Reoptimizing as time progresses — shrinking the horizon and updating inputs continuously.
SoC State of Charge. The energy currently stored in the battery (MWh or %).
SoH State of Health. Remaining capacity as percentage of original nameplate capacity.
Stochastic optimization Optimizing expected value across multiple scenarios with assigned probabilities.
Terminal value Value assigned to energy remaining at end of optimization horizon. Prevents myopic emptying of battery.
TSO Transmission System Operator. Manages the high-voltage grid and procures balancing services (e.g., TenneT, 50Hertz, Amprion, TransnetBW in Germany).
η (eta) Round-trip efficiency. Fraction of energy recovered after a full charge-discharge cycle. Typical: 0.85–0.92 for Li-ion.

Rechtliches: Impressum