A chiller retrofit is often written into a green-factory application as “replaced with high-efficiency centrifugal, saving 150,000 kWh/year”. The number may be correct and still be rejected, because the evidence chain is incomplete: no separate metering, no weather correction, no measured post-retrofit run.
Green-factory evaluation rewards verified energy performance. The retrofit has value at four specific points, and each point needs a different document. Understanding that mapping is more useful than repeating that the plant is efficient.
1. Where the retrofit counts
| Evaluation point | What is assessed | What the retrofit provides |
|---|---|---|
| Unit product energy | Whether the site is below the industry advanced value | Lower cooling electricity per unit of output |
| Equipment efficiency | Whether major energy equipment reaches grade 2 or better | New nameplate, label and test report |
| Energy-saving project | Whether there is a real retrofit investment | Contract, invoice, acceptance report |
| Metering and management | Whether cooling is separately metered and monitored | Chiller-plant meters and data platform |
There is also an indirect benefit: lower cooling power reduces the site’s carbon account, which matters for export customers and carbon disclosure. It does not by itself earn a point, but it supports the energy and carbon narrative.
2. The baseline
The baseline should be at least 12 consecutive months of separate cooling electricity, cooling output or load, and operating hours. A monthly total electricity bill is not a cooling baseline. If the site has no separate meter, the baseline can be estimated from run hours and a measured input power, but the estimate is weaker and should be labeled as such.
Record the chiller model, the refrigerant, the nominal capacity, the control sequence and the cooling-tower condition. A baseline without the plant configuration cannot be corrected later.
3. Weather and load correction
Comparing two years directly is unsafe: a mild summer lowers the cooling load and makes the retrofit look better; a hot summer hides the saving. The common correction is degree-day normalization:
E_corrected = E_measured × CDD_baseline ÷ CDD_report
Use daily weather-station data for the site, and state the base temperature. If the production output changed, correct for output as well. The method should be chosen before the retrofit, not after the numbers are known.
4. A worked example
Take a 1,200 kW chiller plant with 3,000 equivalent full-load hours and 3,600,000 kWh of annual cooling. The old plant has an IPLV of 4.8; the new plant has 6.0.
| Item | Before | After | Difference |
|---|---|---|---|
| Annual cooling | 3,600,000 kWh | 3,600,000 kWh | — |
| Annual electricity | 750,000 kWh | 600,000 kWh | −150,000 kWh |
| Annual cost at 0.65 CNY/kWh | 487,500 CNY | 390,000 CNY | −97,500 CNY |
| Coal equivalent at 0.1229 kgce/kWh | 92.2 tce | 73.7 tce | −18.4 tce |
| CO₂ at 0.58 t/MWh | 435 t | 348 t | −87 t |
The arithmetic is clean, and the simple payback is less clean. At a 1.8 million CNY investment, 97,500 CNY/year gives an 18.5-year simple payback. On electricity savings alone, the project does not justify itself.
5. Why do it anyway?
Three reasons make the project rational even when the energy payback is long. The old chiller may be at the end of its life, with failures and refrigerant leaks; replacement is then a necessity, and the efficient model is a choice within it. The old refrigerant may be phased out, making replacement a compliance issue. The site may have an energy-intensity target, a carbon quota or a customer audit that values the verified saving beyond the tariff.
In a green-factory application, the retrofit’s value is the verified reduction and the evidence, not the simple payback. State the driver explicitly; an evaluator can accept a compliance-driven project with a long payback, but not an investment claimed to pay back in three years when the arithmetic says eighteen.
6. Do not stop at the chiller
Replacing the chiller alone often delivers only about 60% of the expected saving. The remaining waste sits in the pumps, the cooling tower and the staging:
Fixed-speed pumps. A 1,200 kW plant may have 60–90 kW of pumping. If the pumps run at fixed speed, their share rises after the chiller becomes more efficient.
Cooling-tower fans. All fans running at low load, or cycling on a fixed sequence, waste energy in the shoulder season.
Manual staging. Operators start an extra machine to be safe, leaving two machines at low load instead of one at a good load.
A package of chiller, pump VSD, tower control and group control typically delivers 1.5–2 times the saving of the chiller alone, at less than half again the investment. The application should describe the boundary; a chiller-only claim invites the question of why the rest of the plant was ignored.
7. The evidence chain
Before: 12 months of separate cooling energy, chiller nameplate and label, operating log, retrofit plan and savings estimate.
During: contract with the guaranteed efficiency, delivery and installation records, refrigerant recovery certificate, old-unit disposal record.
After: 12 months of operation, third-party efficiency test at the project condition, measurement and verification report, invoices and payment records.
Throughout: calibration certificates for the meters. A meter without calibration undermines the data, however complete the spreadsheet.
Common questions
Does a high-efficiency chiller automatically earn green-factory points?
No. The evaluation looks for a verified reduction in unit product energy, a compliant equipment grade and a real project. A nameplate and an invoice are not enough.
Is third-party verification required?
For government programs and incentive claims, usually yes. For internal management, a self-measured saving is acceptable if the method and data are retained.
What if there is no separate cooling meter?
Estimate from run hours and measured power, label the estimate, and install separate metering during the retrofit so the next project has a baseline.
How much saving is lost if only the chiller is replaced?
Often around 40% of the system potential, because pumps, towers and staging remain inefficient. The exact share depends on the plant.
Is refrigerant recovery part of the evidence?
Yes. It demonstrates compliant disposal and supports the environmental section. Keep the recovery record with the project file.
What is the most common evidence failure?
A savings claim without a 12-month baseline, weather correction or measured post-retrofit data. The number may be right, but it cannot be verified.
What evidence counts?
Metered savings, a baseline and a measurement plan. A design claim is not evidence. The green factory reviewer wants the measurement boundary.
What pays?
Load reduction, part-load control and heat recovery. A chiller replacement alone may not pay. Combine measures.
Further technical reading
- Chiller Retrofit Payback
- GB 19577-2024 Chiller Efficiency
- Chiller Plant Group Control
- Chiller Remote Monitoring
LONGCOM can define the retrofit boundary, the baseline and the measurement plan, and check the savings arithmetic before it enters a green-factory application. For a retrofit evidence review, contact 400-878-0280 or 13002001280.
Related reading
Chiller Retrofit Payback · Chiller Tender Technical Response · Chiller Acceptance Standards