The concepts that make the Certified Energy Manager (CEM) material demanding are the ones that involve judgment: matching an economic metric to the decision it serves, pairing a retrofit with the load shape it actually faces, and writing recommendations whose assumptions survive review. Prepare by rehearsing the full chain from raw facility data to a ranked, documented recommendation rather than memorizing formulas in isolation. Work case scenarios, score them against a rubric, and treat administrative details — scheduling, eligibility, exam format — as the domain of AEE, the credential issuer.
Converting Field Data Into Comparable Energy Units
An energy manager's first technical step is expressing every input in one consistent unit system before comparing fuels, systems, or sites. Practice converting between kWh, therms, MMBtu, and ton-hours until the arithmetic is automatic.
Mixed units are where comparisons quietly break. A gas bill may report therms, a fuel oil delivery in gallons, electricity in kWh, and chilled water in ton-hours, each implying a different heat content. A head-to-head comparison of cost per unit of delivered heat only works after conversion to a common basis such as MMBtu. Build a one-page conversion card, use it in every practice problem until you no longer hesitate, and label each figure as site or source energy.
Worked example: a boiler burns 1,000 therms of gas, about 100 MMBtu of heat. A vendor quotes an electric option at 44,000 kWh for the same duty — 440 kWh per MMBtu. Electric resistance heating needs roughly 293 kWh per MMBtu (1 kWh is about 3,412 Btu), so the quote implies losses or extra load beyond even a resistance benchmark. The conversion did more than change units; it flagged an implausible input, so question the quote's efficiency assumptions before any cost comparison.
- 1 therm is about 100,000 Btu, or roughly 29.3 kWh of heat content
- 1 MMBtu = 10 therms — a fast mental anchor for gas work
- kW/ton and COP both express chiller efficiency but are not interchangeable; convert before ranking options
- Always label whether a figure is site energy or source energy before comparing fuels
Choosing the Economic Metric That Should Decide a Retrofit
Simple payback, return on investment, net present value, and savings-to-investment ratio answer different questions. The skill to build is matching the metric to the decision: short-term screening, capital budgeting, or comparing alternatives.
Scenario one: choose between two lighting retrofits. Option A costs $10,000 and saves $4,000 per year; simple payback is 2.5 years. Option B costs $18,000, saves $3,500 per year in energy plus roughly $800 per year in avoided lamp replacements, and lasts longer. Screened on payback alone, Option A wins, and the maintenance savings and longer life of Option B never enter the calculation. Applying a screening tool to a long-lived capital decision is the trap built into this scenario.
The better decision is to compute life-cycle cost or a savings-to-investment ratio for both options over a common analysis period, including energy savings, maintenance savings, measure life, and the time value of money. Option B's longer service life and recurring maintenance stream can reverse the ranking — and that matters, because a facility adopting Option A faces a second retrofit and repeat installation disruption far sooner. Match the metric to the decision's horizon, and treat maintenance savings as part of the measure.
| Metric | Question it answers | Best used for | Blind spot |
|---|---|---|---|
| Simple payback | How fast is the cash recovered? | Quick screening of small measures | Ignores measure life, time value of money, and post-payback savings |
| Return on investment | What annual return does the measure earn? | Communicating value to management | Distorted when savings are uneven year to year |
| Net present value | What is the net worth of the project in today's dollars? | Ranking independent projects under a budget | Needs a defensible discount rate |
| Savings-to-investment ratio (SIR) | How many dollars of present-value savings per dollar invested? | Comparing long-lived alternatives | Can flatter measures with long tails of small savings |
Separating Demand Charges From Energy Consumption on the Bill
A kWh reduction does not automatically cut the demand charge, and a kW reduction does not automatically cut consumption cost. Bill analysis requires reading each cost component separately before attributing savings to a measure.
Scenario two: two production shifts produce nearly identical monthly consumption, yet one site pays a substantially larger bill because every major motor starts at the beginning of a shift, creating a tall, brief peak. A colleague proposes an occupancy-based lighting retrofit and projects savings proportional to total kWh. Occupancy lighting savings occur during occupied hours and barely dent a peak set by coincident motor starts, so the projected bill reduction is overstated.
The better approach is to decompose the bill first: energy charges follow consumption, demand charges follow the highest kW interval, and power-factor penalties follow reactive behavior. Only then match a measure to the component it actually moves. Staggering equipment starts, trimming short coincident peaks, and correcting poor power factor attack demand-side dollars; occupancy controls and equipment efficiency attack consumption. In case practice, writing 'which cost line does this measure touch?' above every recommendation is a fast, repeatable discipline.
Matching HVAC and Motor Measures to the Actual Load Shape
Efficiency measures only pay when they match how the equipment actually runs. A variable-speed drive helps loads that spend most hours throttled or partially loaded; it does little for a pump or fan running near full output around the clock.
Scenario three: a plant engineer sees a rebate for variable-frequency drives and proposes one for a cooling water pump that runs at full speed continuously because the system was sized for peak summer flow year-round. Assuming 'VFD equals savings' without checking the duty cycle is the plausible mistake. If the pump never throttles and no bypass or excess head exists to recover, the drive adds conversion losses and cost with no flow reduction to harvest.
The better decision is to log the load first: operating hours, percentage of full flow over a representative period, and how flow is controlled. For a fan or pump that spends most hours at reduced output, the cubic relationship between speed and power for centrifugal loads means even modest speed reductions yield large power reductions — but that holds only for variable-torque loads with genuine part-load hours, so treat it as conditional. For the constant-speed pump, better candidates are rebalancing, impeller trimming, or resizing. Fixed drill: for every measure, write the load shape — hours, percent load, control method — before naming it.
Writing an Audit Recommendation That Survives Scrutiny
A defensible recommendation states the baseline, the measurement basis, the assumptions, and the interactive effects between systems. Vague savings claims and undocumented assumptions are the characteristic weakness of audit-style answers.
Strong audit documentation answers four questions a reviewer will ask: What is the baseline and how was it established? How will savings be verified after implementation? Which assumptions — rates, hours, efficiencies — drive the result? And how do systems interact? A lighting retrofit in a conditioned space reduces cooling load but adds winter heating load; omitting that interaction overstates savings, because the heating penalty partially offsets the cooling benefit and the net figure is the honest one.
Training this habit is straightforward. For every measure you analyze, write one sentence on the baseline, one on measurement and verification intent, one on the dominant assumption and what happens if it is off by a factor of two, and one on cross-system effects. You will notice that operating hours and energy rates usually dominate the estimate, which tells you exactly where a sensitivity check belongs. This reflex carries into professional practice, where M&V planning separates funded recommendations from stalled ones.
A Two-Case Practice Routine With a Self-Check Rubric
Convert study time into case practice: take one mock utility bill and one measure list, then work the full chain from data to ranked recommendation. Score yourself against a rubric instead of only checking the arithmetic.
Exercise setup: write or find a practice case with a monthly electric bill (consumption, demand, rate), a gas bill in therms, and three proposed measures — an occupancy-based lighting retrofit, a compressed-air leak repair, and a motor replacement. Convert everything to a common unit, decompose the bill into energy and demand components, estimate savings for each measure, rank them by an appropriate metric, and note one interactive effect and one dominant assumption for each.
Expected observations: if you cannot say which cost line each measure touches, you are ranking on intuition; if your lighting savings do not change when you halve the assumed operating hours, you never identified hours as the dominant assumption; if you ranked a long-lived option by simple payback alone, revisit the metric table above. Score each case out of five points — conversions, bill decomposition, metric matched to decision, assumptions and interactions documented, and a clear ranked recommendation. A consistent four-plus is a learning milestone, not a score prediction.
- Rubric line 1: all inputs converted and labeled (site/source, unit, rate basis)
- Rubric line 2: energy, demand, and power-factor cost lines addressed separately
- Rubric line 3: economic metric justified against the decision horizon
- Rubric line 4: dominant assumption named, with a sensitivity comment
- Rubric line 5: interactive effects between HVAC, lighting, and process noted
A Domain-Ordered Preparation Sequence and Readiness Checks
Sequence your study the way a case unfolds: core energy concepts first, then measurement and interpretation, then applied decision-making, then documentation and professional standards. Reserve the final phase for timed case analysis across all domains.
A realistic adaptable sequence: spend the first block on energy fundamentals — units, combustion basics, electrical demand, HVAC and motor behavior — because every later topic depends on them. The second block covers assessment and interpretation: reading bills, building energy accounts, and identifying load profiles. The third applies that foundation to decisions, including fuel choices and the economics from the metric table. The fourth covers methods, documentation, and professional and ethical standards. Finish with repeated mixed-domain case practice, because that is how the material actually arrives.
Readiness checks you can self-administer: convert any bill figure to MMBtu without notes; given two measures, name which metric should decide and why in two sentences; decompose a bill into energy and demand components and predict which line a measure affects; state a baseline, a verification intent, and a dominant assumption for any recommendation in under a minute; and articulate where ethical and safety boundaries apply to measurement and savings claims. When you can do all five on unseen cases, shift remaining time to mixed timed practice. For practice questions and review materials, see the site's free practice set and study guide library, and rely on AEE for all administrative details of the credential itself.
- Block 1: core concepts — units, fuels, electrical demand, HVAC and motor loads
- Block 2: assessment and interpretation — bills, energy accounting, load profiles
- Block 3: applied decision-making — economics, environmental trade-offs, measure selection
- Block 4: methods, documentation, ethics, and professional standards
- Block 5: mixed, timed case analysis with the self-check rubric
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.