8 of 8 monthly peaks called. $18,352 avoided in 8 months.
How a 515,000 sq ft Class A office tower in Boston turned ISO New England (ISO-NE) (Independent System Operator – New England) settlement charges into a controllable monthly line item.
Bill-validated. Not modeled.
First 8 months since go-live, every figure tied to actual settlement data.
Predicted operational window included the actual validated settlement hour every monitored month.
SiteIQ's forecast aligned with the validated settlement peak interval in every monitored month since go-live.
Validated reduction in Regional Network Load during the settlement hour across successful response months.
Bill-derived transmission-style cost avoided, using illustrative RNS rate of $12.40/kW-month.
Reducing load during the right hour — not many hot days.
A large Boston office tower deployed coincident peak forecasting to improve the timing and confidence of its ISO New England coincident peak response strategy. The objective was not simply to forecast hot days, but to identify the monthly settlement hour that materially influenced Regional Network Service (RNS) cost exposure.
In the first 8 months since go-live, SiteIQ's forecast window aligned with the validated monthly settlement peak interval in every monitored month. During successful response months, the building reduced average settlement-hour Regional Network Load (RNL) by 185 kW, lowering estimated RNS-related cost by an average of $2,294 per month.
The result was a more disciplined forecast-to-action workflow, clearer post-event validation against actual settlement logic, and an estimated $18,352 in avoided transmission-style cost over the first 8 months of live operation.
Predicted window. Actual settlement hour. Validated reduction.
One representative month from the 8-month deployment. The forecast window contained the actual settlement hour, and the response captured 185 kW of RNL reduction.
Representative deployment data. Actual settlement hour and load curves vary monthly; this visualization summarizes the pattern observed across 8 of 8 monitored months.
Confidence in the hour, not the day.
Broad understanding. Limited confidence.
Before deployment, the facilities team had a broad understanding of monthly peak-risk conditions but limited confidence in the exact hour when action would matter financially. That uncertainty created two linked risks: failing to respond during the true settlement hour and over-responding on afternoons that ultimately did not set charges.
- No disciplined workflow to move from early warning to day-ahead preparation to final action.
- No rigorous post-event method to prove whether the response aligned with the actual settlement hour.
Cost concentrated in a single hour each month.
In ISO-NE, Regional Network Service exposure is tied to the customer's Regional Network Load during the monthly settlement peak hour. For this customer, the average pre-response monthly RNL was 1,460 kW.
An operational decision layer, not a dashboard.
SiteIQ was deployed as a staged coincident peak workflow designed to improve both timing and execution discipline.
Weekly outlook
Directional outlook on monthly peak-risk conditions.
Day-ahead
Confirmation when a candidate hour became actionable.
4-hour and 2-hour
Updates refining the likely settlement window.
Final approach
Short-horizon guidance as the peak hour approached.
Post-event ✓
Validation using billing and interval meter data.
Measured operational changes — no tenant impact.
A conservative response playbook designed to reduce load without materially affecting tenant comfort or lease-critical operations.
Pre-cooled zones
Selected common and perimeter zones pre-cooled before the forecasted settlement interval.
Widened deadbands
Temporarily widened non-critical temperature deadbands during the response hour.
Reduced ventilation
Common-area ventilation reduced where operationally acceptable.
Delayed runtime
Selected discretionary equipment runtime delayed until after the risk window.
Across successful response months, these actions produced settlement-hour load reductions ranging from 140 kW to 225 kW, with an average validated RNL reduction of 185 kW.
Eight months. Eight peaks. No misses.
Forecast accuracy and execution discipline measured month by month.
SiteIQ's forecast window aligned with the monthly settlement peak interval in every monitored month.
The predicted operational window included the actual validated settlement hour in each monitored month.
The building executed a full response in seven months and was constrained in one month by tenant event activity.
Realized savings depended on execution as well as signal quality.
Before and after, on the bill.
Using the customer's average validated RNL performance and an illustrative RNS rate of $12.40/kW-month.
Validation methodology
A major strength of the engagement was disciplined post-event validation. In ISO-NE, performance can be tied to Regional Network Load, which reflects the customer's hourly integrated load during the settlement peak hour for the relevant transmission network. This means validation is grounded in settlement logic rather than in a single short-duration spike.
- 01Collect the utility invoice pages showing delivery or transmission-related parameters.
- 02Identify RNL directly or derive the equivalent settlement load basis from the bill.
- 03Convert interval meter data to hourly integrated load where required.
- 04Match the bill-derived settlement load to the hour in the customer load profile.
- 05Confirm whether SiteIQ's predicted window included that hour.
- 06Estimate avoided cost by comparing baseline RNL with validated post-response RNL.
From hot-day reaction to monthly cost control.
By month eight, the building had shifted from an uncertain hot-day response mindset to a more disciplined monthly cost-control process. Forecasting became operationally meaningful because the signal was connected to a repeatable playbook and a clear validation method.
The customer gained clearer confidence about when response was economically justified, a stronger coordination process between facilities and cost management, and a more defensible post-event savings story tied directly to billing logic.
For this Boston office tower, SiteIQ created value by helping the customer reduce load during the right hour rather than reacting broadly across many hot days. The outcome was not just a better forecast — it was a more structured operating process, a more rigorous validation methodology, and a more credible financial story tied directly to ISO-NE settlement logic.
Coincident Peak Forecasting
Identify the hour. Act on it. Prove it worked.