Prepare for the Certified Energy Auditor exam by working numeric audit scenarios across the published body of knowledge: normalize utility data, select the right audit level and economic metric, estimate system-level savings with interactive effects, and document the reasoning. Use the worked examples, comparison table, and rubric below as your core practice loop.
Turning twelve months of bills into a normalized baseline
Energy use analysis starts with a weather- and operation-normalized baseline, not raw bills. Practice converting billing kWh into energy use intensity, then adjusting for heating degree days, cooling degree days, and production or occupancy changes.
Build one worked baseline from a fictional twelve-month bill set before anything else. Compute annual kWh, divide by conditioned floor area to get EUI, and compare monthly consumption against monthly degree days to see which end uses appear weather-driven. Label every intermediate figure. The habit matters because the practice scenarios in this plan include a distractor, such as a month with a partial billing period, and your normalized baseline only reveals it if you compute systematically.
Then separate the concepts that look interchangeable: energy (kWh) versus demand (kW), and consumption versus intensity. A facility can cut peak demand through scheduling while annual kWh barely moves, and an EUI comparison is only valid between buildings of similar use type. Drill unit discipline deliberately: convert kW to kWh over an operating schedule, convert therms or MMBtu into a common energy unit, and state your assumption about billing-period boundaries each time.
- Baseline artifacts to produce per practice scenario: annual kWh and fuel in one common unit, EUI, a demand profile note, and one written normalization assumption.
- Self-check: if you cannot say why the highest-bill month is high (weather, schedule, or rate), the baseline work is not finished.
Matching audit effort to ASHRAE 211 audit levels
The AEE body of knowledge expects you to develop an audit strategy and plan, and ASHRAE Standard 211-2018 gives the commercial audit framework. Practice naming the level of effort a scenario calls for before estimating any savings.
The standard distinguishes a preliminary energy use analysis, a Level 1 walk-through survey with a brief report, a Level 2 energy survey and analysis with end-use breakdowns and ECM identification, and a Level 3 detailed analysis of capital-intensive modifications. The classification drills in this plan describe client requests in words, and the error to train against is jumping to Level 3 detail when the task described is a Level 2 screening, or recommending a walk-through when the deliverable stated is an investment-grade analysis.
Work this as a classification drill. Write one-paragraph facility scenarios, then answer three questions: What deliverable did the client request? What data exists versus what must be measured? What depth of ECM costing is justified? The AEE materials note that CEAs work under ASHRAE 211 and to guides such as the DOE Guide to Energy Audits, so treat scope discipline as subject matter in its own right, not paperwork. A recommendation that exceeds the contracted scope is a wrong answer even when the engineering inside it is sound.
Choosing the economic metric before ranking ECMs
Economic analysis is a decision about which metric fits the question. Simple payback screens speed; life-cycle cost and net present value compare alternatives; internal rate of return and savings-to-investment ratio support capital prioritization.
Scenario one, worked example: a client must choose between retrofit A, costing $40,000 and saving $10,000 per year for 8 years, and retrofit B, costing $65,000 and saving $12,500 per year for 20 years. Simple payback favors A (4.0 years versus 5.2 years). But computing life-cycle cost or net present value at conventional discount rates used in audits, roughly in the single digits to mid-teens, favors B, because A stops saving after year 8 while B persists. The plausible mistake is ranking both measures on payback alone; the better decision is to match the metric to the decision type, first cost screening versus long-lived capital comparison. It matters because an investment-grade recommendation justified by the wrong metric understates a 20-year measure and can steer the client to the smaller project.
Train the distinctions explicitly rather than as a formula pile. Payback ignores time value and anything after payback; NPV and LCC discount every cash flow and compare on total cost or net benefit; IRR expresses the yield of the investment itself; SIR divides discounted savings by discounted cost and is handy for ranking independent measures under a fixed budget. Note that discount rate matters: pushed high enough, the long stream of B's savings is discounted so heavily that A's shorter, larger cash flows can win, so always state the rate before comparing. In practice sets, write one sentence per measure naming the metric you used and why it fits. That sentence is the reasoning the scenario is actually testing.
| Metric | What it captures | Best use in an audit |
|---|---|---|
| Simple payback | Years to recover first cost; ignores time value and later savings | Quick screening of low-cost measures |
| Net present value / life-cycle cost | All discounted cash flows over the analysis period | Comparing alternatives with different lifetimes |
| Internal rate of return | Discount rate at which the project breaks even | Communicating investment yield to owners |
| Savings-to-investment ratio | Discounted savings divided by discounted cost | Ranking independent measures under a budget cap |
Estimating lighting and motor savings without double counting
Lighting and motor/drive savings interact with building loads. Reducing lighting power cuts internal heat gains, which saves cooling energy but adds heating energy. VFD savings on centrifugal fans and pumps follow pump and fan affinity relationships, not a flat percentage.
Train the interactive-effect chain explicitly. If a lighting retrofit removes 10 kW of connected load operating 4,000 hours per year, the direct savings are 40,000 kWh, but the net figure must add cooling savings from the smaller heat gain and subtract the extra heating energy in a heating-dominated season. Practice calculations that supply a cooling-to-heating split check whether you apply both corrections in the correct direction. Write the sign convention down once: less internal gain lowers cooling energy (a savings) and raises heating energy (a penalty).
For motors, distinguish motor efficiency upgrades from speed control. Replacing a motor raises efficiency at similar output; a VFD on a centrifugal fan or pump changes flow and the required input power falls roughly with the cube of speed under affinity relationships, subject to system and static-pressure realities. The recurring error is applying the cube relationship to constant-pressure or static-head-dominated systems, or applying any speed-control logic to equipment that does not vary flow. Always state the load type, duty cycle, and load factor before computing anything.
Reading BAS trends versus taking spot measurements
Data collection questions ask you to match the measurement method to the variable and its variability. BAS trend logs capture schedules and diversity over time; spot metering captures a momentary condition; a one-time reading cannot characterize an intermittent load.
Practice classifying data needs: schedules and setpoints come from the BAS or EMCS where trending exists; electrical demand profiles need interval data at a resolution fine enough to show the operating pattern; combustion efficiency, airflow, and water temperatures need field measurement, and a single reading is only valid if the load is steady. The scenario error to train against is recommending a spot measurement for something that cycles, such as a rooftop unit with staged compressors, where a trend over a representative week is the honest data source.
Work a documentation habit into this section. For each data element, record source, date, instrument or log interval, and operating conditions, because the audit report must let a reviewer reconstruct how savings were estimated. In practice sets, take any scenario paragraph and list which numbers are measured, which are assumed, and which come from the utility or BAS. Anything left unclassified is an assumption you must declare. This classification is a learnable skill and improves fastest with deliberate reps on mixed scenarios.
Worked compressed-air scenario: catch the savings overstatement
Compressed-air scenarios test load profile reasoning. Leak repair, pressure reduction, and storage interact, and speed-control savings logic only applies to compressors that can actually modulate flow efficiently.
Scenario two, worked example: a plant runs a 75 kW lubricated screw compressor, loaded 60 percent and unloaded 40 percent of an 8-hour shift. A night-shift leak survey suggests leaks represent roughly 25 percent of average compressed-air demand. The plausible mistake is claiming leak repair saves 25 percent of the motor's full 75 kW times all operating hours, which ignores that unloaded rotary-screw machines still draw substantial power at idle and that the savings apply to the leak's share of compressed air produced, not nameplate input. The better decision is to start from a demand profile, estimate average compressed-air output, size the leak load within it, and apply the compressor's actual part-load power behavior before converting to kWh. It matters because the inflated figure is exactly the kind that fails an investment-grade review.
Layer a second decision on the same scenario: the client also proposes lowering header pressure from 125 to 110 psig and adding storage to stabilize events. Train the judgment that these measures interact with leak repair, since lower pressure reduces leakage flow and demand-side reductions change the load profile the compressor sees. Sequence the analysis as leak load first, then pressure, then supply-side control, and note in the report which savings depend on which measure. That dependency statement is what separates a defensible recommendation from a stack of overlapping claims.
A two-week practice sequence with readiness checks
Sequence preparation by body-of-knowledge area, one numeric scenario set per day, and finish each set with documentation written as if a reviewer will read it. Readiness means you can produce a defensible baseline, metric choice, and savings estimate for any unfamiliar scenario.
A realistic adaptable sequence: days one and two, baselines and unit drills on billing data; days three and four, audit levels and scoping classification; days five and six, economic metrics on paired-measure scenarios; days seven and eight, lighting, HVAC, and DHW savings with interactive effects; days nine and ten, motors, drives, and compressed air; day eleven, envelope and controls (BAS, EMCS) reasoning; day twelve, alternative generation and storage plus transport and process loads; days thirteen and fourteen, mixed timed scenario sets with full write-ups. Adjust the weighting toward whichever area produced the weakest rubric scores.
Practical exercise with expected observations: take one month of utility bills from any building you can access legitimately, or a constructed dataset, and in one hour produce EUI, a weather comment, the three largest suspected end uses, and one candidate ECM with a payback calculation. Self-check rubric: two points for a correct EUI and normalization assumption, two for a stated data source for each end-use guess, two for an ECM whose metric matches the decision type, and two for naming one interactive effect. Scoring 6 or higher on two different buildings indicates you are applying the material, not reciting it; treat the score as a learning milestone, not a prediction of any exam result. Administrative details such as eligibility and scheduling belong to the certifying body, so confirm those directly with AEE rather than through study materials.
Concrete readiness checks before exam day: you can convert between kW, kWh, and common fuel units without notes; you can state the ASHRAE 211 level of effort for a described scope in one sentence; you can compute simple payback and explain when it is the wrong metric; and you can write a savings estimate with its assumptions in under ten minutes. Any check that fails points you to the specific day in the sequence above, which keeps the final review targeted instead of general.
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.