Treat every practice scenario as a scoping decision, not a quiz question. First ask what the client's goal is and what data exists. Then choose the assessment depth, name the limits of the evidence, and let the recommendations match that depth. This habit, practiced on paper scenarios and the exercise below, is the through-line of the topics listed for BEAP preparation.
The three-level framework: why assessment depth is the first decision you make
ASHRAE's energy audit framework describes three ascending levels of assessment. Level 1 is a walkthrough with preliminary analysis, Level 2 adds an energy survey and analysis, and Level 3 involves detailed analysis with measurements. Choosing among them is a scoping skill you can drill.
Each level exists because client goals differ. A portfolio owner screening many buildings needs a quick identification of candidate measures; a single owner committing capital needs defensible, measurement-backed savings estimates. The levels are a cost-versus-confidence ladder: each step up requires more data, more time, and more rigor, and produces findings you can rely on for larger decisions.
In practice, the level drives everything downstream: which documents you request, which analysis is legitimate, and how firmly you can state recommendations. When studying, read each scenario twice before answering: once for the client's intent, once for the data inventory. If those two facts are clear, the appropriate level usually follows, and so do the limits on what you can claim in the report.
| Attribute | Level 1 (walkthrough) | Level 2 (energy survey & analysis) | Level 3 (detailed analysis) |
|---|---|---|---|
| Primary purpose | Screening; identify candidate measures and low-cost opportunities | Break down energy use; evaluate measures with reasonable estimates | Support major capital decisions with rigorous analysis |
| Data typically involved | Utility bills, brief site observation, basic building information | Extended utility analysis, system inventories, operating schedules | Short-term or long-term measurement, detailed modeling or metering |
| Output | Preliminary findings and recommended next steps | Prioritized measures with cost and savings estimates | Measurement-supported savings and implementation detail |
| Best fit when | The client wants direction, not commitments | The client is comparing measures for planning | The client is committing to a specific investment |
Benchmarking and EUI: normalizing before you compare anything
Energy use intensity (EUI) expresses consumption per unit of floor area, giving buildings of different sizes a common scale. But raw EUI misleads when buildings differ in use, climate exposure, or operating hours, so normalization and comparability checks come before any conclusion.
Compute EUI by dividing annual energy use by floor area, and convert fuels to a common unit before combining them. The trap is comparison: an office and a laboratory, or a building with a data room and one without, are not comparable on EUI alone. Occupancy schedules, process loads, and plug loads can legitimately explain differences that a careless assessor would attribute to inefficiency.
When you benchmark, therefore, do three things in order: verify the floor area, group the building against a genuinely similar peer set, and note which load categories are process-driven rather than HVAC-driven. Practice this by computing EUI for sample datasets and writing one sentence on why each comparison is or is not fair. That sentence-writing habit is what turns a calculation into assessment judgment.
Utility data analysis: separating baseline, weather response, and schedules
A utility analysis should decompose consumption into a baseline load and a weather-sensitive component. Monthly bills reveal this shape; the analytical mistake to avoid is treating every variation in the bills as an efficiency problem needing a measure.
Plot monthly consumption and look for the pattern: a flat floor suggests baseline loads such as ventilation, lighting during occupied hours, and equipment that runs continuously, while seasonal swings track heating and cooling. A building whose consumption barely changes across seasons has a dominant baseline; that observation points your walkthrough toward schedules, controls, and plug loads rather than the chiller.
Weather normalization adjusts consumption for how hot or cold a period was, so you can compare year over year fairly. Its limit matters for scoping: a single year of data gives a weak basis for normalization, and gaps or estimated readings weaken it further. In study scenarios, state explicitly how many billing periods exist and whether they are actual reads, because that determines whether a Level 2 style analysis is supportable or whether you should recommend more data first.
Scenario one: the client who wants a full audit from thin data
A plausible scenario: a property manager asks for a complete energy audit of a 1990s office building and hands over twelve utility bills, with no verified floor area, no equipment inventory, and no operating schedules. The scoping decision is the exam-relevant decision.
The tempting mistake is to accept the full-audit framing and start producing equipment replacement recommendations with cost and savings figures. That is overreach: without verified floor area, benchmarking is unreliable; without schedules, savings from controls measures cannot be estimated credibly. A report built this way makes confident-sounding claims the evidence cannot carry, and the client may commit capital on them.
The better decision is a Level 1 scope: conduct a walkthrough, use the bills for a preliminary load shape, present candidate measures as a ranked shortlist for further study, and state in the report which data gaps block deeper analysis. Then recommend a Level 2 with a defined data request. This matters because the deliverable now matches its evidence, and the client gets an honest, actionable path instead of fragile numbers. Practice this pattern by listing, for any scenario, three data items that would change your recommended scope.
From findings to measures: interactive effects and why packages cannot be summed
Energy conservation measures interact. Reducing lighting energy also reduces the cooling load but can increase heating demand, and schedule changes alter the savings of everything else. Assessment quality shows in how you account for these couplings and sequence low-risk measures.
Interactive effects mean individual measure savings are not additive. Lighting retrofits cut internal gains, so cooling savings shrink or grow depending on the measure set, while heating systems work harder in cold climates. Building envelope and controls measures shift loads onto other systems. A careful assessor evaluates measures as a package, or at minimum flags which figures were computed independently and could shift once others are implemented.
Sequencing is the practical companion. Operational and scheduling changes are typically low cost and reversible, so they come first; their effect on baseline loads should then be reflected in how subsequent capital measures are evaluated. When you study a scenario with three or four measures, write out the interaction chain for each pair before estimating package savings. Naming which measures the savings interact with demonstrates the interpretive skill the assessment topics are aimed at.
Scenario two: the school package where the savings were double counted
Second scenario: an assessor evaluates a school with aging chiller, old fluorescent lighting, and equipment left running overnight. Three measures are proposed, their savings are summed, and the package looks compellingly short on payback. The error is structural, not arithmetic.
The mistake is that the lighting retrofit savings were calculated assuming the existing chiller, and the chiller replacement savings were calculated assuming the existing lighting. Once both are installed, the cooling load is lower than either calculation assumed, so the summed total overstates package savings. The overnight schedule change reduces runtime hours for everything, which changes the annual operating hours underlying the other two estimates as well.
The better decision is to restate the package as a set of ordered steps: implement the schedule correction first at minimal cost, recompute the lighting retrofit against the revised baseline, and evaluate the chiller replacement last against the load that actually remains. Present each figure with its assumptions labeled. This matters because a school board allocating a capital budget on the summed figure would face a shortfall, whereas the sequenced presentation gives them a defensible investment case. Rehearse this by recomputing a two-measure package both ways in a labeled worked example and observing the gap.
A preparation sequence, classification drill, and readiness checks
Prepare in four passes: framework first, benchmarking and utility analysis second, measures and interactions third, then scenario synthesis. Pair every pass with the classification drill below so the scoping decision stays at the center of your practice.
A four-week adaptable sequence: week one, internalize the three levels and write a one-line justification for each level choice in five mini-scenarios. Week two, practice EUI computation, peer-set selection, and load-shape reasoning on sample utility data. Week three, work measure evaluation: interactions, sequencing, and assumption labeling, using the two scenarios above as templates to rewrite with different building types. Week four, synthesize full paper assessments and self-score them. Shift the proportions to match the domains you find weakest; the order, not the calendar, is the point.
The classification drill: write ten two-line scenarios mixing client goals (screening, planning, capital commitment) and data conditions (bills only, bills plus walkthrough access, bills plus metering). For each, name the level, the missing data, and one claim you could not make at that level. Self-check rubric, with four points as the milestone: two points if level and missing data are both correct, one if your unsupported-claim statement truly exceeds the scope, one if you named a data item that would upgrade the scope. Anything below three signals you should revisit sections one and four.
Readiness checks before you sit: you can state the three levels and the decision criteria between them without notes; you can compute an EUI and explain two reasons a comparison might be unfair; you can explain interactive effects in one sentence and give a concrete example; you can outline what an assessment report must disclose about its evidence and limitations. For administrative matters such as current eligibility, scheduling, and exam availability, confirm directly with ASHRAE, since this guide covers learning content rather than logistics.
- Level classification: can you justify a Level 1, 2, or 3 choice from client goal plus data inventory alone?
- Evidence limits: can you name, for any claim, the data that would and would not support it?
- Interaction reasoning: can you trace a two-measure interaction chain and say why summed savings overstate the package?
- Documentation discipline: can you list what a report must state about assumptions, normalization limits, and data gaps?
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.