Reading the window columns: the engines split window gain differently: E+'s "Fenestration Conduction" includes inward flowing absorbed solar (all of it booked as "instant"; about 80% of that column here); RIME's is pure UA·ΔT. Win Total (Solar + Conduction) is the like for like comparison, with one more step on Levels 2 and 3. E+'s pulse decomposition pushes transmitted solar through decay curves built from a pulse distributed by surface area, but beam solar actually lands on the floor; the storage the decay curve misses is not dropped, it appears as negative Interzone Ceiling / Wall / Floor rows (the adiabatic surfaces still holding the afternoon's solar at the peak). RIME's superposition attribution nets that storage against the window itself. So the fair E+ window figure is Win Total E+ plus its interzone rows: L2 Conference B: RIME 10,183 vs E+ window rows 10,784 (-6%), E+ interzone rows -779, reconciled 10,005 (+1.8%); L3 Conference B: RIME 10,947 vs E+ window rows 11,860 (-8%), E+ interzone rows -1,514, reconciled 10,346 (+5.8%). The same input envelope only twin (wall + window + the same adiabatic mass, no internal gains) puts the two engines at +2.4% at the Level 2 peak and +0.4% at Level 3, which is the physics check behind that reconciliation. Level 1 runs the other way for the slab reason (item 4): E+'s ground coupled floor soaks up the beam solar (large negative Ground Contact Floor, small Fenestration Solar) and RIME's shallow ground sink is smaller. Roof cooling column: roof ENERGY agrees between the engines (same U/area/α; inside face conduction within 1% on the day with the DOE-2 film and the face correction); the cooling column still differs by peak hour attribution (E+ reports inside face conduction including absorbed room radiant, RIME the sol-air-driven share), with the difference absorbed by the other columns; zone totals agree. An attribution split, not missing energy. Roof heating row: RIME’s design heating loss is the steady UA·ΔT of the sizing convention; E+’s includes the roof radiating to a clear design night sky, so its roof row reads higher. A convention, and RIME’s building heating total is still above E+.

Result: sum of room sensible cooling peaks RIME +2.4% vs EnergyPlus 25.1; design heating +2.1% (net of the slab convention, item 4, the heating gap narrows further). Interior rooms agree within +0.3% in aggregate across all 42 of them, the widest single room being L3 Men at +0.9%, and interior design heating is zero in both engines. Heating components: exterior wall within 3.54% of E+ (29,232 vs 30,306 Btu/h), roof within 15.8% (19,380 vs 23,020 Btu/h); upper-floor window totals within single digits on all four orientations; peak hours coincident. The only rooms outside ±15% are the Level 1 perimeter spaces, all traceable to the ground-sink convention (item 5). Every residual difference below is a named convention; all act in the conservative direction for RIME.

1. Inputs verified identical, room by room (59 rooms, zero discrepancies; RIME-vs-EPlus-Inputs.html): occupancy by seat count at 450 Btu/h·person (250 sensible / 200 latent); lighting and equipment W/sf per A-601 in every space type; 75/70°F setpoints; infiltration 0.05 cfm/sf of gross exterior wall in both engines; wall assembly calibrated to the governing U-0.064 (builder U-0.0634, R-13 batt with the ASHRAE 90.1 A9.2B steel-stud correction); interior floors 3" LWC on deck per F-1; roof U-0.032 α 0.55; W1 U-0.36 / SHGC 0.38.
2. Scope: commercial kitchen NIC (modeled at 75°F, excluded) and elevator shafts (unconditioned, zero loads) per G-001/A-601. Room loads exclude ventilation outdoor air; design heating takes no internal-gain credit; loads at full 24/7 intensity.
3. Convention alignment. Three EnergyPlus settings are set to what RIME's heat balance documents about itself, so the comparison measures the engine rather than the inputs: interior convection set to Simple (an exact match: RIME's inside film coefficients are the ASHRAE constant film coefficients RIME uses, 0.542 / 0.167 / 0.712 Btu/h·ft²·°F, which E+’s Simple algorithm also carries for wall / ceiling / floor); conduction stated as ConductionTransferFunction (already the E+ default, so no numeric change); timestep 6, since RIME averages six subhourly samples to each hour value. Interior solar absorptances are set to RIME’s 0.50 ceiling / 0.60 floor. Exterior absorptances are NOT changed: the drawings govern them (EW-1 α 0.70, R-1 α 0.55) and both engines already carry those values. The exterior film is the same published model in both engines: the DOE-2 exterior convection correlation (Walton natural convection by tilt and direction of heat flow, MoWiTT wind driven convection at the surface’s height, surface roughness), which RIME computes and the reference uses as its outside algorithm, at the station’s ASHRAE design wind (mean coincident wind speed and prevailing direction with the 0.4% dry-bulb: 4.1 m/s from 350°, carried on every design day of the reference and synced onto the RIME site from the station table). Both engines radiate longwave explicitly to a Clark and Allen sky temperature (emissivity from the design dew point, view split between sky, ground and air by tilt), and both evaluate the film and the sky loss at the outside FACE temperature, which on a massive wall sits far below the sol-air temperature because most of a solar pulse goes into storage. Same-input check: RIME’s outside face is within 1 °F of E+’s on the brick wall and the roof at every hour, and the reference anchor’s walls agree on day conduction within 0.1 to 2% with E+ running its own film unpinned. A Building option, ASHRAE Sol-Air Longwave (Conservative), puts ASHRAE’s tabulated sol-air constants and combined h₀ = 3.0 film back and reads a few percent higher on sunlit surfaces; it is off for the published figures.
4. Reference roof insulation carries its mass (2026-08-27). The E+ model had idealised the R-30 polyiso as Material:NoMass. A real board (6 in at 2.7 lb/ft³) has a small capacitance sitting in the middle of R-30, which gives the roof a lag of roughly an hour; RIME resolves that mass, the NoMass reference could not, and the difference read as a RIME roof lag. All RLVB IDFs now carry the board with its mass; the reference moved −0.7% on building cooling (Level 3 −1.5%), heating unchanged.
5. Roof basis correction (2026-08-27). The RIME model’s roof construction had been entered at solar absorptance 0.45 (RIME’s saved construction default) over R-26.5 of board, where the drawings specify α 0.55 over R-30, the values the E+ reference carries. Found by lining the Level 3 rooms up by roof share of load: the shortfall was a straight line (rooms with only a roof −4 to −6%, rooms where the roof dominates −1 to −3%) and a same input twin reproduced the model’s roof number to 2 Btu/h from those two inputs. Corrected in the model; every Level 3 room now sits within ±1.6%.
6. Fenestration basis correction (2026-08-26). The RIME model’s glazing had been entered as a custom low-e stack that computes to U 0.341 / SHGC 0.46, while the E+ reference carried the plan set’s W-1 at U 0.36 / SHGC 0.38. That is an input mismatch, not an engine difference: on identical inputs, an envelope only twin of Conference B (Level 2) put the two engines at +1.8% of each other at the same hour, and the per-room partition-mass effect agreed within a point. The project glazing was reset to the plan values and the comparison rerun; the earlier +7.6% cooling figure belongs to the mismatched basis. Every Level 2 room now sits within ±0.8%.
7. Glass inward flowing fraction correction (2026-08-27). RIME’s simple glazing optics (the LBNL single layer model both engines use to represent a rated U and SHGC) carried the rated inward flowing fraction of absorbed solar, N = (R_o + ½R_lw) / (R_o + R_lw + R_i), at the NFRC rating films (h_o ≈ 29 W/m²K, winter) and held it fixed; E+ recomputes that split every step at its simulation films. On a design afternoon (h_o ≈ 17, and the inside film the glass actually has) N is about 13% higher, so the glass delivered a tenth less of its absorbed solar than it should and rooms whose peak sits on the falling side of the solar curve (the rooftop floor) read ~2% low. Diagnosed on the same input twin with the interior mass removed from both engines: transmitted solar identical to 0%, the deficit a flat level error in the instantaneous path. The split is now made at the films the glass sees; the interior absorptance defaults were aligned at the same time (partitions and steel deck ceilings 0.50, concrete floors 0.60). Effect: Level 3 −2.4% → −1.9%, Level 2 rooms from ±1.6% to ±0.8%, building cooling +1.1% → +1.5%; heating unchanged. The earlier +1.1% belongs to the rated-film split.
8. Interior partition mass, in both engines. Both models carry the same gypsum partitions (5/8″ gyp / stud-cavity air / 5/8″ gyp) on every interior boundary, as adiabatic mirror mass. Sensitivity experiment (2026-08-26): adding these partitions to an otherwise identical model moved EnergyPlus by +18,125 Btu/h and RIME by +18,063 Btu/h; the two engines agree on the partition physics to 0.3% of the effect. Both move upward, because light gypsum intercepts diffuse solar and radiant gains that would otherwise sink into the heavier concrete floors and returns them sooner. An earlier published figure of +5.3% compared a RIME run without partition mass against an E+ reference with it; that inconsistency flattered the total by about two points and is corrected here.
9. All interior surfaces are adiabatic in BOTH engines for this comparison: the standard single zone loads convention (E+ reference = RLVB-adiabatic.idf: 408 interzone wall, floor and ceiling faces set Adiabatic). Adiabatic surfaces retain their thermal mass in E+; they store and return heat to their own zone, exactly as RIME's single zone floor/partition mass does; only the net zone-to-zone transfer (which neither convention designs to) is removed. Sensitivity: against the realistic multizone E+ run (all interzone surfaces conducting), RIME reads +3.0% cooling / +2.7% heating overall, and E+ interior rooms read 16–25% higher (≤1.1 kBtu/h absolute per room) because neighbor partition surfaces run warmer than room air and elevator lobbies border the unconditioned shafts, differences a single zone design method deliberately excludes.
10. Slab on grade heating uses the ASHRAE F-factor design method, Q = F·P· (T_in − T_{design air}), the standard sizing convention. E+'s GroundFCfactorMethod drives the slab to winter soil temperature (~46°F) and reports about one-third the loss. RIME is conservative and faithful to the method.
11. Ground coupled slab cooling (Building option, default on) uses a monthly shallow ground temperature proxy validated within ~3°F of the SeaTac EPW 0.5 m series in the cooling months. The resulting sink is smaller than E+'s (colder soil + slab solar storage), so Level 1 glazed rooms read +30–45% on window totals, conservative.
12. Exterior film: both engines use the DOE-2 exterior convection correlation at the station's design wind and radiate explicitly to a Clark and Allen sky, evaluated at the outside face; the reference's walls and roof agree with RIME on day conduction within 0.1 to 2%. The Building option ASHRAE Sol-Air Longwave (Conservative) restores the tabulated constants on the combined film. The per-surface "Wall" figures are two different attributions (E+ reports the total inside face conduction including absorbed room radiant, RIME the sol-air-driven share), so compare zone totals, not that column.
13. Infiltration heating is sensible only at standard air density; E+'s component includes cold air density and infiltration latent (humidification), reading ~50% higher on that line.
14. Window gain bookkeeping: RIME reports full SHGC solar + pure UA·ΔT conduction; E+ books the inward flowing absorbed fraction in conduction. Compare Solar + Conduction (Win Total), and on the upper floors add E+'s negative Interzone rows to its side, which is where its pulse decomposition parks the beam solar storage that RIME nets against the window (see the note under the components table). Reconciled that way the Level 2 / 3 Conference B windows agree within 6%.