Every facilities manager or sustainability director who wants to implement building energy optimization eventually has to answer the same question from their CFO or property owner: what is the payback period? And that question is fair — because the building automation and energy management space has a long history of projections that evaporate on contact with reality.
We've been through enough pilots now to have a framework for building an honest ROI case — one that a financially literate person will not dismiss. The framework has three components: establish a defensible baseline, isolate the demand-charge component specifically, and build a year-one projection that accounts for the variables that actually vary. This piece walks through each one.
Why Most Energy ROI Analyses Fall Apart
The typical vendor-produced ROI projection takes your annual energy spend and applies a percentage savings claim. "30% savings on $400,000 annual energy costs = $120,000 per year." A competent CFO will immediately ask: 30% of what, exactly? Gross energy cost? Just HVAC? Weather-adjusted? Compared to what baseline year?
When those questions can't be answered cleanly, the projection falls apart — not because the savings aren't real, but because the methodology is vague enough to be unverifiable. The ROI case doesn't need to be conservative; it needs to be reproducible. If someone else ran the same numbers on your utility bills, they should reach the same baseline figure you're using.
The other common failure mode: bundling energy savings and demand charge savings into a single "energy cost reduction" number without distinguishing between them. This matters because demand charges and consumption charges respond to different interventions. Pre-cooling to avoid peak demand windows has almost no effect on monthly kWh consumption — but it can cut your demand charge in half. If you're modeling a pre-conditioning strategy, building the ROI case around consumption reduction is the wrong metric entirely.
Step 1: Build a Weather-Normalized Baseline
Your baseline is the energy cost your building would have spent without any optimization intervention, adjusted for weather. Weather normalization is not optional — if you run a pilot in April and compare to the previous January, the apparent savings will be heavily contaminated by seasonal load differences. Any CFO who has seen a facilities budget before will notice this.
The standard approach for commercial building baselines is degree-day normalization. You pull your utility billing data for the trailing 12-24 months, pair it with heating degree-days (HDD) and cooling degree-days (CDD) from a local weather station for those same billing periods, and regress monthly energy cost against degree-day totals. The resulting model lets you estimate what your building "should have" spent in any given month based purely on weather, and you compare actuals against that model.
For a 90,000 sq ft office building in Portland, a typical regression might show a base load of around $6,800/month (the portion of energy cost that doesn't vary with weather — lights, plug loads, elevator, server room) plus approximately $42/month per CDD and $28/month per HDD. That regression gives you a defensible counterfactual: "in a month with 280 CDD, this building should have spent approximately $18,600 absent any change in operations."
ASHRAE Guideline 14 (Measurement and Verification of Energy, Water, and Demand Savings) is the appropriate reference standard for this type of baseline methodology. If you want your ROI analysis to survive scrutiny from a sustainability consultant or an ENERGY STAR submission, the Guideline 14 approach for Option B or Option C M&V (measurement and verification) is the right frame. We're not suggesting you run a full ISO 50001 energy review — but citing the methodology gives your analysis credibility with anyone who has done this before.
Step 2: Isolate the Demand Charge Component
This is the step most energy ROI analyses skip, and it's the most important one if you're evaluating a pre-conditioning or peak demand avoidance strategy.
Pull your utility invoices for the past 12 months and separate the bill into its components: energy charges (per-kWh consumption), demand charges (based on your peak 15-minute interval in the billing month), and any fixed charges or distribution fees. For most commercial buildings on standard PGE Schedule 32 or Pacific Power large commercial rates, demand charges represent somewhere between 25% and 40% of the total bill. Some buildings with particularly spiky demand profiles run higher.
The demand charge line item is the number your pre-conditioning strategy is targeting. If your building spent $42,000 last year on demand charges (roughly $3,500/month average), and a credible pre-cooling strategy can reduce the peak demand interval by 20-30%, you're talking about $8,400–$12,600 in annual savings from that single component. That's a precise, defensible number tied directly to a specific line item on your utility bill.
We're not saying demand charge reduction is the only savings mechanism — consumption improvements from reduced HVAC overcycling contribute too. But demand charges are the component with the clearest cause-and-effect relationship to the intervention. Model it separately.
Step 3: Build the Year-One Projection Honestly
A year-one projection for a building energy optimization project needs to account for four things: the savings mechanism, the operating cost of the optimization software, the learning and ramp-up period, and uncertainty ranges.
On savings mechanism: if your baseline analysis shows $38,000/year in demand charges and your pre-conditioning strategy captures 70% of eligible peak demand windows (a realistic figure for a well-tuned system after a learning period), the direct demand charge savings is approximately $26,600/year. Add a secondary benefit from reduced overcycling — conservatively 10-15% of your consumption charge — and you have a total savings range that is grounded in actual bill components.
On operating cost: at our pricing, a 90,000 sq ft building runs $290/month ($3,480/year) on the Single plan. That cost is fixed and known. Net savings year one, assuming a 3-month ramp-up to full optimization, would be approximately $17,500–$22,000 after software costs. Payback on implementation time (the 60-90 minutes of BMS setup) is immediate.
On the learning period: the first 30 days of any predictive control deployment are a learning and calibration period, not a full-performance period. A fair year-one projection should apply roughly 60-70% of steady-state savings to the first three months and full performance thereafter. This is a minor adjustment, but it keeps your projection honest with people who have seen "first month savings" claims that disappear in month two.
On uncertainty: express your projection as a range, not a single number. "Based on our baseline analysis and the demand charge isolation, we project $19,000–$26,000 in annual savings, with the range reflecting weather variability and ramp-up timing." A range with a clear explanation of the bounds is more credible than a single figure — it signals that you've thought about the variables, not that you ran a calculator once.
What the Finance Team Will Ask
In our experience working with facilities managers who've had to take these analyses to ownership groups or property CFOs, there are three questions that reliably come up:
Is the baseline verified against actual utility bills? The answer needs to be yes, with the actual invoice data available. Not a vendor-provided estimate.
What happens if energy prices change? The short answer is that the ROI improves if rates go up (you're saving a larger percentage of a larger bill) and degrades slightly if rates fall. For most commercial customers in the Pacific Northwest, PGE and Pacific Power have historically increased rates 3-6% annually. The direction of risk is favorable.
What is the downside scenario? The honest answer is that in the worst case — the system runs but peak demand windows are rarely triggered due to unusually mild weather or atypical occupancy patterns — the demand charge savings underperform projections and the net benefit is close to zero. That worst case exists, and acknowledging it is what makes the upside case believable.
A Note on What This Framework Is Not
We're not suggesting this analysis will always produce a compelling ROI case. For some buildings — particularly very small ones where demand charges are minimal, or buildings with highly irregular occupancy that makes thermal modeling difficult — the math genuinely might not pencil out at current pricing. The framework is designed to produce an honest answer either way, not to make every project look attractive. If a building's demand charge exposure is $400/month and we're charging $290/month, the net benefit is marginal and we'll tell you that directly.
What the framework does is separate real opportunity from noise. Buildings where demand charges are $2,000/month or higher, with predictable occupancy patterns and existing BMS infrastructure, consistently show favorable ROI. That's the population this analysis is designed for.
The goal isn't a 90-page energy audit report. It's a one-page analysis with three numbers your CFO hasn't seen before: your actual demand charge spend last year, the peak window exposure that is addressable, and the net savings range after software costs. Get those three numbers right and the conversation gets easier.