A Passing WUFI Passive Model That the Mechanical Engineer Disputes
A 60-unit, six-story multifamily project in Minneapolis (climate zone 6A) is modeling to Phius CORE Performance. The WUFI Passive model reports heating demand 14.6 kWh/m²·yr, cooling demand 6 kWh/m²·yr, and a peak heating load of 19 kW. The mechanical engineer's independent load calculation gives 34 kW for the same building. The modeler used a global thermal bridge allowance of 0.10 W/m²K, default window installation psi-values, and no duct losses because 'the ERVs are inside the envelope.' The design includes 8 in. of exterior mineral wool, triple glazing, and an ERV with 82% sensible recovery. A value-engineering proposal would reduce exterior insulation to 5 in. and switch to double glazing with U-0.30. The project has a fixed certification deadline. The modeler says the building still 'passes' on heating demand.
1. Identify the likely sources of the 15 kW discrepancy and explain which are model-input errors versus legitimate differences in method.
Read the worked answer
The 15 kW gap is too large to dismiss. Likely model-input issues include the global thermal bridge allowance, which may understate repeated structural penetrations at balconies, floor slabs, and window installations; default window psi-values, which may not match the actual bracket and sill details; and the omission of duct losses even though ERV ducts inside the envelope still have fan heat and leakage. Infiltration assumptions also matter: a model at 0.6 ACH50 can underpredict actual leakage if the building is not yet tested. On the mechanical side, the engineer may include pickup, diversity, and safety factors that are appropriate for equipment sizing but not for the Phius peak-load metric. Some difference is legitimate method: the modeler may report a block load at a specific design condition, while the engineer may report a sum of room loads with different internal-gain and setpoint assumptions. The key is to reconcile assumptions one by one, not to average the two numbers. The value-engineering change would make the optimistic model even less defensible if the envelope is degraded without a corresponding mechanical redesign.
2. Recommend a QA and parametric plan before accepting the value-engineering change, including specific WUFI Passive outputs to compare.
Read the worked answer
First, rebuild the base model with detailed thermal-bridge calculations for the balcony slabs, window installation, floor-to-wall junctions, and roof parapets. Replace default psi-values with project-specific values, and add duct losses and fan heat unless the ducts are truly within conditioned space and sealed. Compare base and value-engineering cases on heating demand, cooling demand, peak heating and cooling loads, TEDI, source energy, and overheating hours. Also compare monthly heating and cooling profiles, not just annual totals. Run a sensitivity case with infiltration at 0.6 ACH50 and a second case at the tested or expected value. Require the modeler to explain any change in peak load and to show the assembly U-values and window psi-values used. If the double-glazed option is considered, check condensation risk at the frame and spacer, not just U-value. Do not accept 'still passes' as sufficient: the question is whether the model is accurate enough to size equipment and to support certification. A third-party model review before the deadline is cheaper than a failed post-construction test.
3. What would you tell the owner about certification risk and the fixed deadline?
Read the worked answer
Tell the owner that a passing annual heating demand does not prove the building will perform or that the mechanical system is correctly sized. If the model is optimistic on thermal bridges, window installation, or duct losses, the project may fail at the blower-door or commissioning stage, and the certification deadline may be missed anyway. The value-engineering proposal should not be accepted on the basis of a single passing metric. Require a model review and a conservative design case before the deadline, and hold a schedule contingency for re-testing and possible re-modeling. If the team proceeds with 5 in. of insulation and double glazing, the project should expect higher heating demand and peak load; the mechanical design must be updated accordingly. The owner should also understand that Phius certification is based on verified performance, so a model that is not defensible creates contractual and program risk, not just a technical disagreement.