Singapore climate pilot comparisons: keep workload beside energy use

The problem

A Singapore climate technology practitioner sees a pilot's lower energy total but cannot tell whether the technology changed performance or the operating workload simply fell.

Write the comparison you are being asked to make

Keep the reported result and the proposed conclusion in separate lines. “The meter recorded less electricity” describes one quantity; “the replacement technology improved performance” attributes a change. Establish which question matters to you before judging a graph. Record the two periods, named system, measured quantity, unit and source, together with the delivered service in each period. Leave an unclear period or meter boundary unresolved. A presentation may contain enough information to discuss a total while lacking the operating record needed to explain it. Do not repair that gap by assuming the pilot did the same work as the earlier system or by treating an attractive before-and-after slide as a controlled comparison.

Define the output before making a per-unit number

Ask what one counted output represents and whether the definition stays the same in both periods. Completed service units, operating hours and items processed answer different questions. A count should not silently include different kinds of work or exclude unfinished attempts in one period only. Keep the agreed definition beside the number, and mark whether it came from a record or a presenter’s explanation. If the service requirement changed, describe that difference before comparing rates. Choosing a convenient denominator does not make unlike work equivalent. Singapore’s SME Sustainability Hub discusses a baseline and activity data as organisational measurement context; it does not establish the output definition or validity of any particular pilot comparison.

• SME Sustainability Hub: Stage 2 Measure ↗ · Source checked 20 October 2026

Work through one wholly fictional dual-view example

This arithmetic example describes no real pilot, product or Singapore event. Assume both rows use the same fictional meter boundary and count completed units under the same fictional service definition. Earlier row: 600 kWh and 100 completed units. Pilot row: 480 kWh and 60 completed units. The absolute observation is 120 kWh less electricity. The per-unit values are 600 divided by 100, or 6 kWh per completed unit, and 480 divided by 60, or 8 kWh per completed unit. These statements can coexist: the pilot row has a lower total and a higher measured amount per counted unit. Neither calculation explains why the rows differ. The workload and other operating conditions still need examination before attributing the difference to equipment.

Use a worksheet that preserves both views

Make one row per period with these columns: period; measured system and unit; total recorded use; service definition; completed output; calculated use per output where meaningful; source; and open condition. Retain the original values when adding a calculation, and show the division so another reader can inspect it. If the output is missing or its definition differs, leave the rate blank and explain why. Do not replace the total with the rate or choose whichever produces a favourable headline. The Hub names energy intensity, emissions intensity and absolute emissions separately when discussing progress. That distinction supports asking which view answers the actual decision; it does not approve this worksheet or convert electricity observations into a verified emissions result.

• SME Sustainability Hub: Stage 3 Act ↗ · Source checked 20 October 2026

List changed and unknown conditions explicitly

Use three columns: confirmed same; confirmed different; and not established. Ask about workload mix, operating period, relevant environmental conditions, service quality requirement and what equipment the measurement includes. Record only what the available account establishes. A system that operates fewer hours may show a lower total even when its per-output measure rises; that is a question to investigate, not proof of the cause in your case. The fictional example supplies two descriptive observations and no controlled test. Do not decide that the technology is effective or ineffective merely from that pair. A discussion should identify which difference affects the comparison and what existing information or appropriate further assessment would clarify it.

Prepare an evidence request the presenter can answer

Choose the largest unresolved comparison condition and ask for its basis. For example: “Do both periods count completed service under the same definition, and where is that definition recorded?” Or: “Which operating conditions changed with the pilot, and which were kept comparable?” Ask whether the figures are measured, estimated or extracted from a partial period. You are trying to understand the claim’s scope, not demanding private records at an open event. A simplified public or invented table can make the question clear. If the presenter cannot supply the basis, keep it unresolved instead of filling the cell with a plausible operational assumption. A reply that names a record is a route to evidence, not evidence you have already read.

Choose a session that examines this comparison

Look for an advertised measurement exercise, operating-case discussion or opportunity to inspect how a comparison was defined. A device demonstration may reveal a sequence without explaining two operating periods. An investment panel may discuss commercial direction without examining workload data. Ask what participants inspect and whether the format accepts your permitted table or uses a supplied example. Confirm prerequisites, software and any input rules. Use the broad climate-engineering guide when selecting the system and contribution; use the energy-demo claim guide when the problem is extending electricity observations into a wider environmental statement. Those are separate tasks. A climate topic alone does not confirm that the organiser offers individual pilot analysis.

Keep the conclusion proportional to the observations

Write four lines after the discussion: quantities compared; service definition and conditions established; descriptive conclusion supported; and causal or wider claim still untested. For the fictional rows, you can report lower total electricity and higher electricity per counted unit under the stated arithmetic assumptions. You cannot report a proven equipment benefit, carbon saving or performance failure from those figures alone. Keep a suggested follow-up as a proposal until the authorised evidence is obtained. Use Boardzip’s free climate search for suitable learning candidates, then verify their actual programme and participation terms. If the needed comparison information remains unavailable, retain the limitation. An event recommendation or repeated slide cannot supply the missing operating record.