Prepare for the CEM by drilling scenario calculations, not term lists. Build a one-page sheet of conversion factors and decision rules, work retrofit and bill-analysis problems that force tool selection, and check yourself with a rubric that measures whether you can justify each assumption you make.
Power vs. Energy: Stopping the Most Common Unit Slip
Power (kW) is a rate; energy (kWh) is power multiplied by time. Demand charges price kW, consumption charges price kWh. Confusing the two distorts every downstream savings calculation, so make the distinction a conscious checkpoint before computing anything.
In energy management coursework and simplified exam scenarios, the distinction is mechanical: a 100 kW motor running 4,000 hours consumes 400,000 kWh, and its 100 kW rating is what a demand charge responds to. The trap is that both figures appear in the same problem, and a question about peak demand can look identical to a question about annual consumption. Train yourself to write the units beside every intermediate value before multiplying.
A second layer is unit conversion across systems. Textbook energy problems move between kWh, Btu, therms, and MMBtu, so keep a small conversion card: 1 kWh ≈ 3,412 Btu, 1 therm = 100,000 Btu, 1 MMBtu = 10 therms. In a worked example, converting a boiler fuel input in therms to a common Btu basis is what lets you compare it against an electric alternative. Without the conversion habit, two correct-looking answers compete and only one is dimensionally sound.
- kW → instantaneous demand; kWh → energy over an interval
- Demand charges bill the peak kW; consumption charges bill total kWh
- Core conversions: 1 kWh ≈ 3,412 Btu; 1 therm = 100,000 Btu
- Self-check: label units on every line of your scratch work
Rate Structures in Practice: When Demand Beats Consumption
Utility bills in exam scenarios typically combine a consumption charge, a demand charge, and sometimes a power factor clause. The decision rule: load-shifting measures save demand dollars only if they cut the billing peak, while efficiency measures save consumption regardless of when they run.
Work a paper scenario: a plant bills at $0.08/kWh plus $12/kW of monthly peak demand. A compressed-air project eliminates 30 kW at all hours. If it runs during every peak interval, the demand saving is 30 × $12 = $360/month, plus the consumption saving of 30 kW × operating hours × $0.08. A plausible mistake is pricing the 30 kW once and ignoring its hours, or pricing only the kWh and forgetting the demand line entirely. The better decision is to compute both components separately and state that the demand saving requires the reduction to appear at the billed peak.
This is also where load factor earns its name. Load factor is average demand divided by peak demand over the period; it describes how flat your load profile is, not how much energy you used. A high load factor at the same consumption means a lower peak and therefore lower demand charges. When a scenario offers a scheduling change that flattens a profile without reducing kWh, the correct analysis shows zero consumption savings but real demand savings. Candidates who reach for a consumption formula here will defend the wrong project.
Worked Retrofit Scenario: A Lighting Swap Done Two Ways
Retrofit questions ask you to separate demand reduction (kW) from annual energy reduction (kWh) before pricing anything. The error worth practicing against is applying the wrong time base to one of the two, which roughly doubles the apparent benefit in this scenario.
Scenario: 100 fixtures convert from 400 W metal halide to 150 W LED, operating 4,000 hours per year. Correct structure: demand reduction = 100 × (400 − 150) / 1000 = 25 kW. Energy reduction = 25 kW × 4,000 h = 100,000 kWh per year. The plausible mistake is multiplying the 25 kW by 8,760 hours as if the load ran continuously, yielding 219,000 kWh and an overstated payback. The better decision is to anchor hours to the fixture schedule the problem states, then price kWh at the consumption rate and 25 kW at the demand rate if the fixtures operate at the billing peak.
Why the distinction matters: the two errors lead to different project rankings. At $0.08/kWh the corrected energy saving is about $8,000 per year; the erroneous figure suggests $17,500, which could flip a marginal project from reject to fund. Add the mechanism, not just the arithmetic: many instructors note that reduced lighting heat can lower cooling energy in cooling-dominated buildings, but that credit depends on climate and system type, so in a simplified scenario state it as a conditional term rather than folding it in silently. Explicit assumptions are what a reviewer can audit.
- Step 1: compute kW reduction from fixture wattage difference
- Step 2: multiply by stated operating hours, never 8,760 by default
- Step 3: price kWh and kW separately under the given rate
- Step 4: note conditional interactions (cooling credit) as stated assumptions
Motors, Pumps, and Fans: Affinity Laws Set Realistic Savings
For variable-flow systems, textbook affinity rules say power scales roughly with the cube of speed (or flow) in an ideal pump or fan. That cubic relationship makes small speed reductions look dramatic, so exam problems test whether you temper it with efficiency and system reality.
In simplified problems, reducing pump or fan speed to 80% of nominal gives ideal power near 0.8³ ≈ 51% of the original. A plausible mistake is reporting a 49% energy saving on the motor input, ignoring that drive losses, motor efficiency at reduced load, and static pressure requirements reduce the realized percentage. The better decision is to present the affinity result as an upper bound, then apply the efficiency adjustments the problem provides. When a scenario gives you a fan law exponent or a system curve hint, that hint overrides the pure cube rule, and using the generic cube anyway is the error the problem is built to reveal.
Connect this to motor replacement decisions: replacing a standard-efficiency motor with a premium-efficiency unit saves the load percentage times the efficiency difference, not the full nameplate difference. A 100 hp motor at 75% load with a two-point efficiency gain saves roughly 75 hp-equivalent × the efficiency delta, converted to kW and hours. Practicing both calculations side by side teaches the selection rule: affinity laws govern flow control projects; efficiency deltas govern equipment swap projects. Mixing the two frameworks is where retrofit comparisons go wrong on paper and in procurement.
HVAC and Envelope: Degree Days and Chiller Metrics
Weather-driven consumption is normalized with heating and cooling degree days; chiller performance is compared with COP, kW/ton, and part-load values. The selection rule: use degree days for envelope and weather questions, and efficiency metrics for equipment comparison questions.
Degree-day normalization converts monthly fuel or electricity use into a per-degree-day intensity so two months with different weather can be compared. In a paper exercise, divide each month's heating energy by that month's heating degree days; a stable ratio suggests weather explains the variation, while a falling ratio suggests a real efficiency change. The mistake to avoid is comparing raw monthly kWh across seasons and concluding consumption rose because equipment degraded, when hotter weather is the stated driver. That reasoning error is exactly what a well-built scenario dangles in front of you.
Chiller metrics need their own decision rule: COP is dimensionless (cooling output divided by power input), kW/ton is its inverse in mixed units (lower is better), and integrated part-load values acknowledge that chillers rarely run at design conditions. When a scenario asks which chiller choice is better, convert both to the same metric before comparing; a trap answer pairs a COP figure against a kW/ton figure and lets you pick the numerically smaller one. State the conversion you performed, because the justification is part of the answer, not decoration around it.
Measurement and Verification: Choosing an Option and Adjusting the Baseline
M&V questions ask which verification approach fits a given project and how the baseline should be adjusted for known changes. The rule: match measurement effort to the saving's size and risk, and separate routine adjustments (weather, occupancy) from non-routine ones.
The commonly taught framework divides M&V into option-style approaches: measure key parameters with stipulated others, measure all parameters of the isolated system, compare whole-facility utility bills against a regression baseline, or use calibrated simulation. A small lighting retrofit with predictable hours fits the first approach; a variable-speed pump project whose savings depend on flow patterns fits the second; a whole-building tune-up with many interacting measures fits whole-facility billing analysis. A plausible mistake is demanding full measurement on every project, which inflates cost, or using bill analysis for a measure too small to see in facility-level noise.
Baseline adjustment is the second skill. Worked scenario: a post-retrofit month shows higher kWh than the baseline year, and the mistake is declaring the measure failed. The better decision is to apply the documented adjustment terms: cooling degree days rose, production volume rose, and the regression baseline predicts higher use for those conditions, so savings are judged against the adjusted baseline, not the raw history. Why it matters: without this separation, weather and production swing the verdict of every energy project, and the savings figure loses credibility with management.
| M&V approach | Best-fit project type | Cost/effort | Main limitation |
|---|---|---|---|
| Key parameter measurement, others stipulated | Predictable retrofits, e.g., lighting with fixed schedules | Low to moderate | Stipulations carry the risk if hours are wrong |
| All-parameter measurement of isolated system | Variable-load equipment, e.g., VFD on a pump | Moderate to high | Requires metering and instrumentation expertise |
| Whole-facility billing analysis | Building-wide projects with many interacting measures | Low instrumentation | Savings must be large enough to rise above noise |
| Calibrated simulation | Complex or not-yet-implemented design cases | High | Model credibility depends on calibration quality |
Economic Metrics: Payback First, Then Life-Cycle Truth
Simple payback is years of annual savings to recover cost; life-cycle cost and net present value compare options across the full horizon with discounting. Use payback for screening, LCC or NPV for choosing among durable alternatives.
Worked example: a measure costs $40,000 and saves $10,000 per year, so simple payback is 4.0 years. The mistake worth practicing against is treating equal paybacks as equal projects. Two measures can both pay back in 4.0 years yet differ in lifetime: one lasts 5 years, the other 20. The better decision is to compute life-cycle cost or NPV when comparing long-lived options, because payback ignores everything after the recovery point and every dollar's timing. State the discount rate a problem gives you; if none is given, in simplified exam terms you compare undiscounted totals and say so.
Build the selection habit with a small drill: for each practice problem, write which metric the question implies. A question about how quickly cash recovers implies payback; a question about which of two chillers to buy over 20 years implies LCC or NPV; a question about rate of return implies comparing annual savings to investment as a percentage. Practicing the identification step separately from the arithmetic is faster than re-deriving it under time pressure, and it makes wrong-tool errors visible in your error log where you can fix them.
- Simple payback = installed cost ÷ annual savings; screening only
- LCC sums all costs discounted over the analysis period; use for alternatives
- NPV discounts future savings; positive NPV supports proceeding
- Drill: label each practice question's implied metric before calculating
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.