Study CEM-AU by making decisions, not reciting formulas. Practise ranking measures under tariffs, screening them with the correct financial metric, and attributing savings to honest baselines. Use the two worked scenarios, the bill-analysis exercise, and the readiness checks below to confirm you can apply each concept rather than merely recite it.
Why kWh Savings Alone Misrank Energy Projects
Energy charges bill total consumption in kWh; demand charges bill the peak power drawn in kW. A measure's value depends on which component it reduces, so ranking projects by raw kWh saved can invert the correct order.
An energy charge multiplies every kilowatt-hour consumed by a rate, so it rewards anything that runs less or runs more efficiently. A demand charge bills the single highest power draw over a billing interval, so it rewards measures that reduce load during the facility's peak, even briefly. These are different products: one prices volume, the other prices capacity the utility must reserve for you.
Apply this by tagging every candidate measure with the tariff component it touches. A scheduling change that trims an evening peak affects demand; a night-shift efficiency upgrade may affect only energy. Two measures with identical kWh savings can differ sharply in dollar savings once you know when the kWh occur and whether the peak kW drop. Build this two-column habit - energy effect, demand effect - before you practise any financial arithmetic.
- Energy charge: kWh consumed × rate - rewards reduced or more efficient operation at any time
- Demand charge: peak kW × rate - rewards load reduction during the billing peak specifically
Worked Scenario: A Lighting Upgrade Against a Compressed-Air Fix
Ranking measures by kWh saved alone is the trap. Under a tariff with demand charges and time-of-use rates, the smaller-kWh measure can deliver more dollars. Work the tariff before ranking.
Worked example (illustrative numbers only). A facility pays a flat 25 c/kWh energy charge, an off-peak rate of 18 c/kWh, and $15 per kW of monthly peak demand. Measure A, a lighting retrofit, cuts 40 kW of on-peak load and saves 4,000 kWh/month. Measure B, a compressed-air leak repair, saves 6,000 kWh/month but runs almost entirely off-peak with no peak-demand reduction. A plausible mistake is to rank B first because 6,000 exceeds 4,000 kWh.
The better decision prices each measure. Measure A: 4,000 on-peak kWh × $0.25 = $1,000, plus 40 kW × $15 = $600 in demand savings, totalling $1,600/month. Measure B: 6,000 off-peak kWh × $0.18 = $1,080, with no demand effect. A outperforms B in dollars despite saving fewer kilowatt-hours. Why it matters: measure ranking drives capital allocation, and a kWh-only ranking points the budget at the weaker project. Rehearse this flip until pricing-before-ranking is automatic.
| Measure | kWh saved/mo | When it occurs | Demand effect | Monthly $ value |
|---|---|---|---|---|
| A: lighting retrofit | 4,000 | On-peak hours | −40 kW at peak | $1,000 + $600 = $1,600 |
| B: compressed-air leak fix | 6,000 | Off-peak hours | None | $1,080 |
Load Factor and Peak Demand: Reading the Bill Before the Retrofit
Load factor - consumption divided by peak demand times hours - shows how evenly a facility uses power. Computing it from a bill tells you whether demand management or consumption efficiency is the richer opportunity.
Load factor compares what you consumed with what you would have consumed running at peak output all period. A high load factor means steady operation, where energy charges dominate and efficiency measures pay. A low load factor means spiky operation, where demand charges and peak-trimming measures deserve first attention. It is a diagnostic, not a savings figure: it steers where to look, never replaces measure-specific arithmetic.
Practise on paper bills before touching candidate measures. Extract monthly kWh, billing demand in kW, and days in the period; compute load factor and state which tariff component a given measure would move. This trains a valuable interpretation habit: reading operating character from billing data rather than jumping straight to equipment lists. Practise stating the conclusion in one sentence - 'low load factor, so peak-trimming first' - as a habit that makes your reasoning defensible before you commit to detailed numbers.
- Inputs from a bill: monthly kWh, billing demand (kW), hours in the period
- Interpretation: high load factor → energy efficiency focus; low load factor → demand and peak management focus
- Guardrail: load factor ranks opportunities directionally; it never prices a specific measure
Simple Payback, ROI, and Life-Cycle Cost: Choosing the Right Screen
Each financial metric answers a different question, and using the wrong one distorts rankings. Match the metric to the decision: quick screening, like-for-like comparison, or long-life total-cost choices.
Simple payback answers how fast the capital returns and is a screening tool, not a decision tool: it ignores everything after payback and the time value of money, so it systematically favours cheap short-lived fixes. Return on investment expresses an annual return rate and suits comparing measures of similar life, but it still says nothing about total cost over a long service life.
Life-cycle cost and net-present-value style measures answer the total question: capital plus operating, maintenance, and energy costs discounted over the asset's life. They suit long-lived, maintenance-heavy choices and discretionary rankings, but they inherit your discount-rate assumption, so always state it. In practice, first ask what decision is being made - screen, compare, or commit - then name the appropriate metric. Writing 'simple payback is 3.2 years, but the asset lasts 20, so life-cycle cost should confirm' demonstrates the distinction cleanly.
| Metric | Question it answers | Use when | Main limitation |
|---|---|---|---|
| Simple payback | How fast does capital return? | Fast screening, small capital | Ignores post-payback savings and time value |
| Annual ROI | What yearly return does it earn? | Comparing similar-life measures | Blind to service-life differences |
| Life-cycle cost | What is total cost over asset life? | Long-lived or O&M-heavy choices | Sensitive to discount-rate assumption |
| NPV-style ranking | What value does it add in today's dollars? | Ranking discretionary projects | Assumptions drive the result |
Baseline Discipline in M&V: Stopping Double-Counted Savings
When multiple measures share one baseline, their effects overlap and interact. Attribute savings measure-by-measure, account for interactions and operating changes, or the reported total will not survive scrutiny.
Worked example (illustrative numbers only). A facility used 1,200,000 kWh in a baseline year. In one period it replaced a chiller and reprogrammed BMS schedules; consumption fell to 1,050,000 kWh. A plausible mistake is to attribute all 150,000 kWh of savings to the chiller because it is the flagship project. This fails two ways: the scheduling upgrade contributed its own savings, and the two measures interact - better schedules change the load profile the new chiller serves, so neither measure's isolated saving simply adds up.
The better decision separates attribution. Estimate each measure's own effect from submetered data or engineering calculations, then reconcile the sum against the metered change, noting the interaction explicitly. Adjust the baseline for drivers that changed independently - weather, production output, occupancy - so you are not crediting a measure for a quieter year. Why it matters: measurement and verification exists to make savings claims credible, and an attribution you cannot defend undermines every other number in the report.
- Define the baseline period and normalise for weather, production, and occupancy changes
- Attribute savings per measure using submetering or defensible engineering estimates
- State interactive effects between measures rather than assuming savings add linearly
Unit Discipline: Conversion Checks That Keep Scenario Arithmetic Honest
Scenario arithmetic is a chain of linked steps where a single unit slip - mixing kWh, MJ, and GJ, or confusing energy with power - silently corrupts everything downstream. A fixed unit-checking routine catches slips before they cascade.
Anchor two conversions: 1 kWh = 3.6 MJ, so 1,000 kWh = 3.6 GJ; and power (kW) versus energy (kWh) differ by time, so a 10 kW load running 5 hours uses 50 kWh. Write units at every step and carry them through multiplication - a result labelled 'kW' that should be 'kWh' exposes the error immediately. Because tariff and savings calculations link several steps, the risk is not any single conversion but the way an early slip becomes invisible once embedded in later arithmetic.
Practical exercise. Take one recent electricity bill, real or constructed, and record monthly kWh, billing demand in kW, and days in the period. Compute load factor, convert monthly consumption to GJ, and write one sentence naming which tariff component each of two candidate measures would affect. Expected observation for a sample month: 120,000 kWh, 400 kW peak, 30 days gives 120,000 ÷ (400 × 720) ≈ 41.7% load factor, and 120,000 kWh ≈ 432 GJ.
- Self-check rubric (score each 0-2, target 8/10):
- Units labelled on every intermediate step, power never confused with energy
- Load factor computed with correct hours in period, within a plausible range
- Conversion kWh↔MJ↔GJ correct (×3.6 scaling)
- Tariff component (energy vs demand) named for each measure
- One-sentence conclusion states the opportunity direction before detailed maths
A Five-Week Preparation Sequence with Readiness Checks
Sequence study from concepts to decisions: core energy concepts, tariff reading, financial screening, baseline and M&V logic, then integrated scenario drills. Close with readiness checks, not just completed pages.
Week 1, master core energy concepts and unit discipline with the rubric exercise until conversions are error-free. Week 2, practise reading tariffs - flat, time-of-use, and demand components - and computing load factor from sample bills. Week 3, drill financial screening: given three measures, name the right metric and rank them, defending the metric choice in one sentence. Week 4, work baseline and M&V scenarios until attribution and interaction effects feel routine rather than exceptional.
Week 5, integrate: full scenarios that combine tariff, finance, and verification reasoning, plus flashcards and mind maps that link each concept to a decision trigger rather than a definition. Readiness checks: you can price two measures under a time-of-use plus demand tariff without ranking errors; you can explain why overlapping measures cannot share one baseline claim uncritically; your unit rubric score reaches 8/10 consistently; and you can justify each financial metric choice in one sentence. Use the free practice questions to test these under time, and revisit weak topics rather than rereading everything.
- Sequence: units and concepts → tariff reading → financial screening → baseline and M&V → integrated drills
- Readiness check 1: rank three measures by dollar value under a TOU + demand tariff
- Readiness check 2: explain double counting and interactions in a two-measure retrofit
- Readiness check 3: self-check rubric score of 8/10 or higher on three consecutive exercises
References and further reading
Use these references to explore the concepts and check the latest information from the relevant organizations.