BPI Multifamily common mistakes
BPI Multifamily · Understand the traps and how to avoid them.
Treating a multifamily building as a single-zone house and applying whole-building pressure or infiltration assumptions uniformly across all units and common areas.
Candidates carry over single-family habits; they forget that compartmentalization, shared walls, and common-area-to-unit pressure boundaries make each dwelling unit its own pressure zone, so uniform assumptions misstate air leakage and driving forces.
What to do instead
Before any pressure or infiltration reasoning, explicitly map the pressure boundaries: each unit, the corridor, stairwells, and mechanical rooms. Reason unit-by-unit and state which boundary a measurement or assumption applies to.
Confusing stack effect, wind effect, and mechanical ventilation pressure in a mid- or high-rise, and attributing observed pressure differences to the wrong driver.
Multifamily buildings are tall enough that stack pressure dominates in cold weather, and exhaust systems add strong negative pressure; candidates underweight height and temperature difference and over-attribute to wind or HVAC alone.
What to do instead
For any pressure scenario, estimate stack pressure from building height and indoor-outdoor temperature difference first, then layer wind and mechanical effects, and state the direction and relative magnitude of each driver.
Selecting a blower-door or compartment test protocol without confirming the unit is prepared (registers taped, exhaust sealed or in intended state, doors to corridor in required position) and without recording those conditions.
Candidates memorize test steps but not the pre-test configuration; multifamily units have many penetrations and shared boundaries, so an unprepared unit yields a number that cannot be compared to a reference or benchmark.
What to do instead
Write out the pre-test checklist for the specific protocol being used, execute it, and record configuration details alongside the result so the measurement is reproducible and comparable.
Reporting a single blower-door result for a multifamily unit without normalizing to the correct reference area or volume, or without noting which reference (conditioned floor area, envelope area, unit volume) was used.
Different programs and benchmarks use different normalization bases; candidates pick whichever they remember and then compare the result to a benchmark computed on a different base.
What to do instead
Always state the normalization basis explicitly and convert or re-derive the metric to match the benchmark's basis before comparing or reporting pass/fail.
Accepting combustion-safety or ventilation measurements taken under non-steady or atypical conditions (doors open, exhaust fans off, appliances at partial load) as representative.
Candidates rush to collect data points without controlling operating conditions, and multifamily shared exhaust makes steady-state hard to achieve, so readings drift and are misinterpreted.
What to do instead
Define and document the operating condition for each measurement (appliance firing stage, fans on/off, doors closed), wait for stabilization, and repeat until readings are stable before recording.
Building an energy model with assumed inputs (R-values, infiltration, HVAC efficiency) that contradict the assessment data actually collected, without flagging the discrepancy.
Candidates treat modeling as a separate exercise from field data and default to typical or code values when a field value is inconvenient or missing, producing savings estimates disconnected from the building.
What to do instead
Tie every major model input to a documented source (field measurement, utility bill, or cited assumption), and where a field value differs from a default, use the field value or explicitly justify the default and mark it for review.
Presenting weather-normalized or modeled savings as a single point estimate without addressing uncertainty, interactive effects, or the baseline definition.
Candidates want a clean number and skip baseline selection and sensitivity; multifamily savings are sensitive to occupancy, base load, and weather year, and interactive effects (e.g., envelope tightening changing ventilation and heating) are ignored.
What to do instead
State the baseline explicitly, quantify or bound uncertainty (range or sensitivity cases), and discuss interactive effects on other end uses before presenting the savings figure.
Writing a scope of work that lists measures without specifying location, quantity, existing condition, access requirements, or who is responsible for what.
Candidates draft scope from a measure list rather than from the assessment findings, so the scope lacks the unit-specific detail a contractor needs and cannot be verified after installation.
What to do instead
For each measure, specify unit or common-area location, quantity, existing condition and measurements, required access and resident coordination, responsible party, and the acceptance criteria that will verify completion.
Prioritizing measures by perceived savings or popularity rather than by assessment findings, health-and-safety urgency, and cost-effectiveness for the specific building.
Candidates default to familiar measures (lighting, aerators) and skip the diagnostic-driven ranking; they also fail to put health-and-safety items (combustion, moisture, mold) ahead of efficiency measures.
What to do instead
Rank measures using a documented sequence: health and safety first, then envelope and air-sealing that enable other measures, then systems, then low-cost items, each justified by assessment data and a stated cost-effectiveness basis.
Producing a report that presents raw data and measure lists without the required standard references, test conditions, assumptions, limitations, and clear findings-to-recommendations linkage.
Candidates treat reporting as a data dump and omit the documentation that makes results defensible; they also fail to connect each recommendation back to the specific finding that motivated it.
What to do instead
Use a report structure that states the standard or protocol followed, test conditions and instrumentation, assumptions and limitations, then presents each finding with its linked recommendation and the verification method.