Certified Energy Manager (CEM)

Issuing body: Association of Energy Engineers (AEE) Jurisdiction: International (US-centric code/standards basis; SI-unit exam versions exist for international candidates) Level: Professional certification, ANSI-accredited to ANSI/ISO/IEC 17024; recognized by the US DOE as a qualified certification (IRS Notice 2023-59) Eligibility: engineering or science degree (or equivalent education/experience combinations — AEE publishes six eligibility paths trading education against years of energy-management experience); pass the CEM exam Renewal: every 3 years — renewal application documenting professional development and continued employment in energy management

Exam facts verified from current AEE materials (aeecenter.org, crawled within days of writing): the computer-based exam is 4 hours, 130 multiple-choice questions, open book, scored out of 1040 points (8 points/question) with a passing score of approximately 700. Sections I and II are mandatory for all candidates — (I) Energy Accounting and Economics and (II) Energy Audits and Instrumentation — and candidates choose 9 of the remaining 13 sections. The 15 sections: Energy Accounting and Economics; Energy Audits and Instrumentation; Electrical Systems; HVAC Systems; Motors and Drives; Industrial Systems; Building Envelope; Combined Heat & Power and Renewable Energy; Building Automation and Control Systems; Thermal Energy Storage; Lighting Systems; Boiler and Steam Systems; Maintenance and Commissioning; Fuel Supply and Pricing; Financing, Performance Contracts, Measurement & Verification. (An older AEE guideline document describes an 80-question version — the current CBT format per AEE's self-evaluation exam page is 130 questions / 4 hours; confirm the current candidate bulletin before sitting.)

1. What the credential is

CEM is the flagship credential for people who reduce energy cost and consumption in buildings and industry — energy managers, ESCO engineers, utility program staff, and consultants. It certifies both the measurement discipline (audits, instrumentation, M&V) and the decision discipline (economics, project justification, financing). The open-book format is deliberate: AEE is testing whether you can do energy management with references at hand, the way the job actually works — which makes it a test of fluency and judgment, not memory.

2. Pathway

Meet an eligibility path (degree + experience combinations; the six paths let experience substitute for formal education at defined ratios).
Prepare with the Body of Knowledge — AEE's primary reference is Guide to Energy Management (Capehart, Turner, Kennedy); the exam is built from it.
Sit the exam: 4 hours, 130 MCQs, open book. Answer Sections I & II plus your best 9 of the remaining 13 — choose sections matching your professional strengths, and mark your choices carefully (only the first 9 marked beyond the mandatory two are scored).
Maintain: 3-year renewal with professional-development and employment documentation.

Exam strategy: the open-book trap and the section-choice game

Open book ≠ easy. 130 questions in 240 minutes = 1.85 min/question. If you look up more than a handful of items, you run out of time. The winning preparation is tabbed-book fluency: know exactly which reference holds which formula (economics tables, motor efficiency tables, lighting coefficients, steam tables) so lookups take 20 seconds, not 3 minutes.
Section choice is strategic. You answer 11 of 15 sections. Drop your 4 weakest entirely — a section you half-know yields ~50% of 8-point questions, worse than a strong section's 90%. Most engineers drop Fuel Supply and Pricing plus two others outside their practice. Decide your 9 before exam day.
Mandatory sections carry the exam. Energy Accounting and Economics + Energy Audits and Instrumentation are everyone's Sections I & II — master them completely, because there is no dropping them.
Memorize: the time-value-of-money formulas and factor notation (P/F, F/P, P/A, A/P, A/F, F/A); energy unit conversions (kWh↔Btu↔therms↔MMBtu); motor and lighting equations below; the M&V option definitions (IPMVP Options A–D). These are high-frequency across sections.
Derive: anything in HVAC psychrometrics from first principles is too slow — instead memorize the decision rules (economizer when outdoor enthalpy < return enthalpy; chilled-water reset raises chiller COP by ~1.5–2% per °F) and look up the chart values.

3. Knowledge domains — theory taught properly

3.1 Energy accounting and economics (MANDATORY)

Energy accounting. Normalize everything: the Energy Use Index (EUI = annual energy / floor area, kBtu/ft²-yr or kWh/m²-yr) and energy cost index ($/ft²-yr) let you benchmark buildings against peers (CBECS data, ENERGY STAR Portfolio Manager scores). Weather-normalize with heating/cooling degree days before claiming savings — a mild winter is not an efficiency project. Submetering and interval data reveal load shapes: baseload (the overnight minimum — pure waste if it's high), peak demand (kW, which drives demand charges), and power factor (utilities penalize low PF; correct with capacitors — know the kVAR calculation).

Time value of money — the full apparatus. Money has time value because of opportunity cost and risk. The six interest factors, all derivable from compound interest F = P(1+i)ⁿ:

Single payment: (F/P,i,n) = (1+i)ⁿ; (P/F,i,n) = 1/(1+i)ⁿ
Uniform series: (P/A,i,n) = [(1+i)ⁿ−1]/[i(1+i)ⁿ]; (A/P,i,n) = [i(1+i)ⁿ]/[(1+i)ⁿ−1] (capital recovery)
Sinking fund: (A/F,i,n) = i/[(1+i)ⁿ−1]; (F/A,i,n) = [(1+i)ⁿ−1]/i Note (A/P) = (A/F) + i — a useful cross-check. NPV = present worth of benefits minus costs; accept if NPV > 0. IRR = the discount rate where NPV = 0; accept if IRR > hurdle rate (MARR). Simple payback = first cost / annual savings — the most-used and most-abused metric: it ignores time value, savings after payback, and project life. Life-cycle cost = first cost + PW(operating, maintenance, replacement, energy) − PW(salvage). The exam's favorite economics trap: a project with 2-year payback but 3-year equipment life vs. a 5-year-payback project with 20-year life — LCC/NPV decides correctly, payback misleads.

Inflation and escalation. Energy prices escalate faster than general inflation — use a real escalation rate for fuel in LCC analyses. Nominal vs. real discount rates: (1 + nominal) = (1 + real)(1 + inflation).

3.2 Energy audits and instrumentation (MANDATORY)

Audit levels (ASHRAE). Level 1 (walk-through): low-cost/no-cost measures, benchmarking, preliminary savings. Level 2 (energy survey and analysis): end-use breakdown, calibrated analysis, evaluated ECOs with costs and savings. Level 3 (detailed analysis): investment-grade — hourly simulation, monitored data, bid-ready ECO definitions. The exam tests matching effort to decision: don't run a Level 3 to justify lighting retrofits.

Instrumentation theory. You must know what each instrument measures and its error sources: power/energy analyzers (true RMS vs. averaging — harmonics fool averaging meters; CT sizing and phase alignment), combustion analyzers (O₂, CO, stack temperature → combustion efficiency via the Siegert formula concept: efficiency falls as excess air and stack temperature rise), psychrometers and data loggers (placement away from heat sources; logging interval vs. memory), ultrasonic flow meters (need full-pipe, developed flow profile — 10 diameters upstream/5 downstream rule of thumb), infrared cameras (emissivity settings — shiny surfaces lie), light meters (measure at the task plane). Calibration: instruments drift; an audit's savings claim is only as good as its measurement uncertainty.

Baselining and adjustment. Measured savings = baseline energy − post-retrofit energy ± adjustments for weather, occupancy, production. This is the conceptual core of M&V (see 3.15) and the exam hammers it: unadjusted comparisons are the most common bogus savings claim.

3.3 Electrical systems; 3.5 Motors and drives

Power fundamentals. P = √3·V·I·PF (3-phase), PF = kW/kVA = cos φ. Demand charges bill the peak kW (often ratcheted — the peak sets 12 months of charges, so one bad 15 minutes costs a year). Load factor = average/max — low load factor means you're paying for capacity you rarely use.

Motors. Induction motor efficiency peaks near 75–100% load and falls off badly below 50% — the oversized-motor disease: a 50 HP motor running a 15 HP load wastes energy continuously. NEMA Premium efficiency tables; the exam expects motor-replacement economics (efficient motor premium vs. operating hours — justified only above ~2,000–4,000 hr/yr). Variable-frequency drives: the affinity laws for centrifugal loads (fans, pumps) — power ∝ speed³. This cubic is the single most lucrative equation in energy management: 20% speed reduction → ~49% power reduction (0.8³ = 0.512). VFDs on constant-torque or static-head-dominated loads do not follow the cube law — know the load type before promising savings (high static head → diminishing returns).

Power factor correction: kVAR_needed = kW × (tan φ_old − tan φ_new). Capacitor placement and harmonic resonance risk (capacitors + transformer inductance can resonate at harmonic frequencies — detuned reactors where VFDs create harmonics).

3.4 HVAC systems

The refrigeration cycle and COP. COP_cooling = Q_evaporator / W_compressor; EER = 3.412 × COP. Carnot sets the ceiling: COP_Carnot = T_cold/(T_hot − T_cold) (absolute temperatures). Real chillers run at a fraction of Carnot — which is why chilled-water temperature reset works: raising CHWS temperature 1°F lifts the evaporator temperature, shrinking the lift (T_hot − T_cold) and improving COP ~1.5–2%/°F. Condenser-water temperature matters more (bigger absolute effect on lift).

Airside. Sensible vs. latent loads; the economizer (free cooling when outdoor air enthalpy is below return — dry-bulb vs. enthalpy control, with enthalpy preferred in humid climates); VAV vs. CAV (VAV with the fan affinity law is the retrofit); simultaneous heating and cooling (reheat — the classic waste; fix with supply-air reset and deadband controls); ventilation per ASHRAE 62.1 (don't "save energy" by violating ventilation rates — the exam's ethics-adjacent trap).

Boiler and steam (3.12). Combustion efficiency vs. thermal efficiency; excess air control (every 1% excess O₂ ≈ associated stack loss — minimize to the CO threshold); steam traps (failed-open traps blow live steam — ultrasonic testing; trap losses are pure waste); condensate return (every 10°F of condensate return ≈ 1% fuel savings); insulation of steam lines (bare flanges and valves are the audit's photo opportunities).

3.6 Industrial systems; 3.8 Building envelope; 3.10 Thermal energy storage; 3.11 Lighting

Compressed air (inside Industrial Systems): the most expensive utility in the plant — ~8 HP of compressor per 1 HP of air delivered (overall efficiency ~10–15%). Every 2 psi of pressure reduction ≈ 1% energy savings; leaks are the measure (a single 1/4" leak at 100 psi wastes ~100+ cfm); fix leaks and lower header pressure before buying a new compressor.

Envelope: conduction (Q = U·A·ΔT), infiltration (the dominant load in leaky buildings — blower-door testing quantifies it), solar gain (SHGC), and thermal mass. Insulation economics follow diminishing returns — the first inch pays, the sixth barely does.

Thermal storage: shift cooling load off-peak with chilled water or ice storage; economics live entirely in the demand-charge and time-of-use rate differential — without a rate differential, storage is just extra equipment.

Lighting: the mature ECO — efficacy (lumens/watt: LED 120–180+ vs. fluorescent ~80–100 vs. HID ~50–120), the lumen method for illuminance (E = (N·n·Φ·CU·LLF)/A), lighting power density limits (ASHRAE 90.1), and controls (occupancy, daylight harvesting — photosensor commissioning is where savings die). Color metrics (CCT, CRI) for occupant acceptance.

3.7 CHP and renewable energy; 3.9 Building automation; 3.13 Maintenance & commissioning; 3.14 Fuel supply & pricing; 3.15 Financing, performance contracts, M&V

CHP: total efficiency = (electric + useful thermal)/fuel input — 70–80% vs. ~35–50% for separate heat and power. Feasibility needs coincident thermal and electric loads (hospitals, campuses, industry); the spark spread (electricity price vs. gas price) is the economic gate.

BAS: control sequences are the measure — optimal start/stop, demand limiting, setpoint reset, scheduling. A BAS with bad sequences wastes more than no BAS (simultaneous heating/cooling on automation is still waste, just scheduled waste).

M&V (IPMVP): Option A (retrofit isolation, key-parameter measurement — stipulate the rest), Option B (retrofit isolation, all-parameter measurement), Option C (whole-facility billing analysis — needs stable baselines, weather normalization), Option D (calibrated simulation — for new construction or complex retrofits). ESPC (energy savings performance contracting): the ESCO guarantees savings and is paid from them; M&V defines the guarantee's measurement — the exam tests matching the option to the project type.

4. Key equations with units and assumptions

Interest factors: (P/F,i,n) = 1/(1+i)ⁿ; (P/A,i,n) = [(1+i)ⁿ−1]/[i(1+i)ⁿ]; (A/P,i,n) = [i(1+i)ⁿ]/[(1+i)ⁿ−1]; (A/F,i,n) = i/[(1+i)ⁿ−1].
NPV = Σ(CF_t/(1+i)ᵗ) − C₀; IRR: NPV = 0; Payback = C₀/annual savings [yr].
3-phase power: P = √3·V·I·cos φ [W]; PF correction kVAR = kW(tan φ₁ − tan φ₂).
Affinity laws (centrifugal): Q₂/Q₁ = N₂/N₁; H₂/H₁ = (N₂/N₁)²; P₂/P₁ = (N₂/N₁)³.
Motor economics: annual cost = HP × 0.746 × hours × $/kWh / η.
COP: Q_L/W; EER = 3.412·COP; Carnot COP = T_C/(T_H − T_C) [K].
Conduction: Q = U·A·ΔT [Btu/h or W].
Lumen method: E = N·n·Φ·CU·LLF/A [lux or fc].
Unit conversions: 1 kWh = 3,412 Btu; 1 therm = 100,000 Btu ≈ 29.3 kWh; 1 MMBtu = 10 therms ≈ 293 kWh; 1 HP = 0.746 kW = 2,545 Btu/h.

5. Worked examples (original)

Example 1 — Motor replacement economics. A 100 HP motor, η_old = 0.91, runs 6,000 hr/yr at $0.10/kWh. Annual cost = 100 × 0.746 × 6,000 × 0.10/0.91 = $49,187. Premium motor η_new = 0.956: cost = $46,820. Savings = $2,367/yr. Premium cost $3,500 → payback 1.48 yr. But note the load: if the motor is 40% loaded, efficiency of both motors drops and the savings shrink — verify loading with a power measurement before promising. This "verify load first" step is the audit discipline the exam rewards.

Example 2 — VFD on a cooling-tower fan. 40 HP fan motor, runs 4,000 hr/yr, can average 80% speed with a VFD. Affinity: power ratio = 0.8³ = 0.512 → saves 48.8% of fan energy. Fan input ≈ 40 × 0.746/0.92 ≈ 32.4 kW; savings = 32.4 × 0.488 × 4,000 × $0.09/kWh = $5,693/yr. VFD cost ~$6,000 → ~1.1-yr payback. Caveat stated on the report: savings assume the tower can actually unload (verify against the chiller's condenser-water temperature requirements — minimum flow limits may cap speed reduction).

Example 3 — NPV vs. payback trap. Project A: $50,000, saves $25,000/yr, 3-yr life (payback 2.0 yr). Project B: $80,000, saves $20,000/yr, 15-yr life (payback 4.0 yr). At i = 8%: NPV_A = 25,000(P/A,8%,3) − 50,000 = 25,000×2.577 − 50,000 = $14,425. NPV_B = 20,000(P/A,8%,15) − 80,000 = 20,000×8.559 − 80,000 = $91,180. Payback picks A; NPV correctly picks B by a wide margin. The exam uses exactly this structure to test whether you know payback's blindness to project life.

Example 4 — Lighting retrofit, lumen method. An office 100 ft × 60 ft needs 50 fc average. LED troffers: 2 lamps... use fixtures at 4,000 lumens each, CU = 0.65, LLF = 0.85. N = (E·A)/(n·Φ·CU·LLF) = (50 × 6,000)/(4,000 × 0.65 × 0.85) = 300,000/2,210 = 136 fixtures. Existing: 200 fluorescent troffers at 96 W each (19.2 kW); new: 136 LED at 36 W (4.9 kW). Demand savings 14.3 kW; energy at 3,000 hr/yr = 42,900 kWh/yr. At $0.11/kWh + $12/kW-mo demand: energy $4,719 + demand 14.3×12×12 = $2,059 → $6,778/yr. Fixture cost 136 × $120 = $16,320 → 2.4-yr payback. Include the HVAC interaction (reduced cooling load) as a stated bonus, not in the base case — exam discipline: don't double-count interactive effects without analysis.

Example 5 — M&V option selection. A chiller replacement in a hospital with stable occupancy: Option B (retrofit isolation, measure kW and ton-hours before/after with installed metering) — the chiller is metered, loads vary with weather so whole-facility billing (Option C) would need heavy normalization. A lighting + controls retrofit across 40 buildings: Option C with weather normalization, or Option A (measure fixture wattages, stipulate hours with sampled logging). New construction LEED claim: Option D calibrated simulation. The exam tests this matching directly.

6. Standards/regulations cross-reference

ASHRAE 90.1 — energy standard for buildings; lighting power densities, envelope, HVAC efficiency minimums. The CEM's design-side reference.
ASHRAE 62.1 — ventilation minimums; the constraint on airside savings.
ASHRAE Guideline 14 — measurement of energy/demand/water savings; the M&V-adjacent technical backbone.
IPMVP — Options A–D; the performance-contracting measurement protocol.
ISO 50001 — energy management systems (Plan-Do-Check-Act for energy); the management-system complement to the CEM's project work.
IEEE 519 — harmonic limits; relevant when VFDs proliferate (total harmonic distortion limits at the point of common coupling).
NEC Article 220/310/430 — load calculations, conductor sizing, motor circuits (for the electrical sections).
Federal ESPC authority / state performance-contracting statutes; utility rebate and demand-response programs; IRA 45Y/48E and 179D — the financing/policy layer (know what each instrument does, not current rates).

7. Pitfalls and trap patterns

Payback worship. Any question pairing a short-payback/short-life option against a longer-payback/longer-life option is testing NPV/LCC — compute, don't eyeball.
Affinity-law overreach. The cube law applies to centrifugal loads with low static head. Positive-displacement pumps, conveyors, and high-static-head systems don't follow it.
Unadjusted baselines. Savings computed from raw before/after bills without weather/occupancy normalization are the exam's canonical wrong answer.
Ventilation violations. Airside "savings" from cutting outdoor air below 62.1 minimums is never the right answer.
Section-marking errors. Marking more than 9 elective sections means only the first 9 count — decide your 9 in advance and mark precisely.
Nominal vs. real rates. Mixing nominal discount rates with real (unescalated) cash flows, or escalating energy costs at general inflation instead of fuel escalation.
Oversized-motor blindness. Always check loading before motor economics — a premium-efficiency motor at 35% load saves little.

8. Practice problems (original, with answers)

P1. A project costs $120,000, saves $30,000/yr for 8 years. At i = 10%, compute NPV and simple payback. Accept or reject? Answer: Payback = 4.0 yr. (P/A,10%,8) = 5.335. NPV = 30,000×5.335 − 120,000 = 160,050 − 120,000 = $40,050 > 0 → accept.

P2. A 25 HP supply fan runs at full speed 5,000 hr/yr. A VFD lets it average 70% speed. Motor input at full speed is 20 kW. Estimate annual kWh savings. Answer: Power ratio = 0.7³ = 0.343; savings fraction = 0.657. kWh saved = 20 × 0.657 × 5,000 = 65,700 kWh/yr. (State assumption: centrifugal fan, low static head.)

P3. Which IPMVP option fits a whole-building retrofit with stable occupancy where utility bills are the only available data, and what adjustment is mandatory? Answer: Option C (whole-facility). Mandatory: weather normalization (degree-day regression) and adjustments for any non-routine changes (occupancy, additions, equipment).

P4. Compute the kVAR of capacitors needed to correct a 500 kW load from 0.75 to 0.95 power factor. Answer: φ₁ = cos⁻¹(0.75) = 41.41°, tan = 0.882; φ₂ = cos⁻¹(0.95) = 18.19°, tan = 0.329. kVAR = 500 × (0.882 − 0.329) = 276.5 kVAR → specify ~275–300 kVAR in switched steps.

P5. Name the two mandatory CEM exam sections and explain the section-choice rule. Answer: I — Energy Accounting and Economics; II — Energy Audits and Instrumentation. Candidates must answer these two plus 9 of the remaining 13 sections (15 total); if more than 9 electives are marked, only the first 9 are scored.

P6. A chiller's COP improves from 4.5 to 5.2 after a controls upgrade. For a constant 300-ton cooling load running 3,000 hr/yr at $0.10/kWh, compute annual savings. Answer: 300 tons = 300 × 3.517 = 1,055 kW thermal. Power before = 1,055/4.5 = 234.5 kW; after = 1,055/5.2 = 202.9 kW. Savings = 31.6 kW × 3,000 h × $0.10 = $9,480/yr.

P7. Why is raising chilled-water supply temperature an energy measure? Quantify the typical effect. Answer: It raises the evaporator temperature, reducing compressor lift (T_cond − T_evap); Carnot COP improves as lift shrinks. Rule of thumb: ~1.5–2% efficiency gain per °F of CHWS reset, limited by dehumidification/latent-load requirements and coil performance.

Suggested study order: Sections I & II to mastery → your 9 chosen elective sections by professional strength → timed open-book practice with tabbed references. The PVIP guide covers the renewables project side; the good-to-know guide covers ISO 50001, ASHRAE 90.1, and the interconnection/policy substrate.

Original ETS reference writing, verified against public sources on 2026-09-30. Thin sourcing is flagged inline. Not legal advice — confirm current editions with the issuing body.

Certified Energy Manager (CEM)