Chiller Test Matrix

Reference engine: EnergyPlus 26.1 Chiller:Electric:EIR with matched CAPFT (Curve:Biquadratic), EIRFT (Curve:Biquadratic), and EIRFPLR (Curve:Quadratic) coefficients drawn from the same production database CogenS uses at run time. Scored against ASHRAE Guideline 14-2023.

Reference engine: EnergyPlus 26.1 N x Chiller:Electric:EIR on parallel branches driven by PlantEquipmentOperation:CoolingLoad with the Optimal load distribution scheme. 21 of 39 in-scope scenarios pass the applicable ASHRAE Guideline 14-2023 band; 6 cases are documented staging-policy Tier-2 divergences where the CogenS equal-load production scheme (mirroring TRANE/JCI/Carrier control sequences) intentionally diverges from EP's Optimal cycle-minimum scheme; 12 PLR=1.00 Scroll/Reciprocating cases are flagged as an open full-load EIRFT evaluation-convention mismatch under follow-on investigation.

Reference engine: EnergyPlus 26.1 two-type chiller plant (n1 x Chiller:Electric:EIR Type 1 + n2 x Type 2 on parallel branches under PlantEquipmentOperation:CoolingLoad Optimal). 4 / 10 scenarios pass ASHRAE Guideline 14-2023 strict on system electric and cooling; 4 / 10 are dispatch-philosophy + EIRFPLR-clamp interactions where CogenS' system-COP-optimized dispatch puts an active unit below the production EIRFPLR validity envelope while EP's Optimal smallest-first keeps it inside (same bankable-TEA-vs-thermodynamic-accounting divergence as Stage 2); 2 / 10 are PLR=1.00 Scroll/Reciprocating cases inheriting the Stage 2 open finding on full-load EIRFT evaluation convention.

Reference engine: EnergyPlus 26.1 WaterHeater:Stratified standalone-run (no plant loop; pure U.A.dT shell physics at 5-node stratification), with the tank held BELOW ambient so the shell term is a heat gain. 6 / 6 in-scope scenarios pass ASHRAE Guideline 14-2023 strict on tank temperature trajectory NMBE (+/- 10%) and CV(RMSE) (<= 30%). RE-SCORED July 2026 against the current stratified-tank engine; the previous engine’s figures are quoted alongside each gate. The shell-gain integrator was previously reported here as carrying a -18% surface-area convention difference with EnergyPlus. It was not a convention difference but an under-count: the engine integrated the cylindrical side wall and not the two end caps, which for this geometry is 8/10 of the surface. The current engine integrates the full surface.

Reference engine: Analytical engineering energy-balance gate on the CogenS engine output (m·cp·ΔT = delivered_charge − delivered_discharge − shell_gain, +/-1% over the full 168-hour cycle) plus a no-overdispatch safety gate (delivered <= requested on both charge and discharge). EnergyPlus cross-validation is DEFERRED — the ThermalStorage:ChilledWater:Stratified isolated-tank EP topology surfaces runtime instabilities (setpoint < tank temperature limit; freeze warnings), documented in the chilled_tes_chargedischarge IDF builder header. The IDF is emitted as a side artifact for future iteration; the validation IS the CogenS energy-balance + dispatch-tracking contract. The Cover sheet, the Reference Output sheet (which shows the REQUESTED charge/discharge schedule), and this matrix row all carry that disclosure so the reviewer is not led to believe a head-to-head EnergyPlus comparison is happening.

Reference engine: Composition self-consistency validation of two EP-validated CogenS component engines: the Stage 1 chiller engine (189/189 vs EnergyPlus Chiller:Electric:EIR) and the Stage 4 chilled-TES shell engine (6/6 vs EnergyPlus WaterHeater:Stratified). Stage 5 verifies that COMPOSING them into either setpoint-buffer or dynamic-deadband coupling reproduces a self-consistent integrated dispatch. Setpoint-buffer mode: energy balance ±0.1%, tank held at setpoint, COP physically sensible — 1/1 PASS regression-free. Dynamic mode (Gap 1 + Gap 2 fixes mirrored from boiler Stage 5 commit 021767ab): P-controller deadband produces oscillating tank-top + sawtooth SOC matching EP IndirectHeatPrimarySetpoint behaviour; 12/12 PASS at relaxed gates (±0.5% balance / 8 C drift) capturing the deadband swing while staying well inside the strict ASHRAE Guideline 14-2023 ±5% annual-energy band.

Reference engine: EnergyPlus LifeCycleCost (NIST Handbook 135 reference implementation) on operating present value, plus closed-form analytical recomputes for NPV / IRR / simple payback / MACRS depreciation. All 12 in-scope scenarios pass strict at 0.0000% on operating PV vs EnergyPlus, 0.0000% on NPV / payback / TCO consistency vs the analytical year-by-year geometric-series expansion, 0.0000 percentage points on IRR vs an independent polynomial-root solver, and $0.00 max absolute on MACRS GDS depreciation vs IRS Publication 946.