Verified Compressor Heat Recovery Projects

Verified Compressor Heat Recovery Projects

01. Executive engineering answer

A verified compressor heat recovery project demonstrates, with dated and traceable measurements, how much useful thermal energy a defined receiving process actually accepted from a compressor heat source. It must identify the compressor and recovery circuit, measure the relevant source and sink conditions, account for simultaneous operating hours and auxiliary consumption, and distinguish useful delivered heat from theoretical availability, exchanger transfer, avoided fuel and financial benefit. The published outcome must have a named source owner, consent, measurement uncertainty and an approved claim boundary.

A compressor dissipates input energy through multiple pathways, but not all dissipated heat is technically accessible, available at a useful temperature, recoverable without compromising the compressor, or demanded by the process. A vendor recovery percentage or motor rating cannot answer this question alone.

02. Ownership and non-overlap

This is an evidence-dependent publication, not an equipment sales claim, engineering design approval or declaration that Reddy Compressors has completed a particular project. A fictional example may illustrate the record but must never be styled as a real client installation.

03. Heat pathways and what may be recovered

Oil-injected rotary screw compressor

Heat may be transferred from the compressor oil circuit through an approved oil-to-water heat exchanger or a purpose-designed recovery package. Preserve minimum oil temperature, cooling control, oil viscosity, pressure drop, thermal protection and manufacturer warranty limits. Recovery cannot override the compressor’s normal cooling duty.

Oil-free screw compressor

Different compression stages, inter/aftercoolers and cooling media can present different thermal levels. Record exactly which stage or cooler supplies the recovery circuit; do not apply an oil-injected recovery percentage to an oil-free package.

Reciprocating compressor

Jacket water, aftercoolers and other approved heat paths may be candidates; intermittent duty, pulsation, temperature and coolant chemistry constrain actual delivery. Treat the manufacturer’s cooling arrangement as the design boundary.

Air-cooled versus water-cooled machines

An air-cooled package may need approved ducting or an engineered heat exchanger; a water-cooled machine may have an accessible liquid loop. Airflow diversion can raise compressor-room temperature or increase fan resistance, so a heat claim cannot omit cooling and ventilation acceptance.

Waste-heat temperature quality

Thermal power at a low temperature is not automatically useful to a higher-temperature sink. Exchanger approach temperature, return temperature, pinch constraints and possible heat-pump input affect feasibility. State source/sink inlet and outlet temperatures, not only a headline kW.

Existing process and seasonal sink

Process preheating, wash water, space heating and boiler-feed preheat have different duty profiles. A sink may be absent at night or in warm months. Record actual acceptance and storage or bypass behaviour, rather than assuming full-year use.

04. Declare four energy boundaries

A B3 value can exceed B4 when hot water circulates through a bypass or when storage loses heat. B5 and B6 are separately evidenced outcomes, not synonyms for B4. Do not double count heat recovered in both hot-water storage and downstream use.

05. Source-side measurement and compressor operating duty

Record exact compressor identity, rated duty and actual loaded/unloaded/variable-speed states during the thermal measurement. Confirm the approved recovery tapping points, fluid type, design pressure, thermal limits, isolation valves, bypass path and the interaction with existing coolers. Include maintenance condition, cooler fouling, ambient temperature, compressor discharge temperature, lubricant type where relevant and recovery-controller setpoints.

For liquid heat recovery, measure volumetric or mass flow and both temperatures at the same timescale; specify sensor insertion depth, mixing, insulation, sensor calibration and time alignment. Source-loop heat extracted may not equal useful sink heat. If the source flow is estimated from pump curves, mark the result estimated and account for valve position and head variations.

06. Sink-side duty, actual usefulness and rejected heat

Document the receiving process, minimum required supply temperature, permitted return temperature, heat demand, operating calendar, process hygiene or separation requirements, thermal storage, existing heating system and standby strategy. A recovery loop that produces hot water when the sink is shut must count that heat as rejected or stored until useful withdrawal is measured. The approved sink owner must confirm whether the delivered temperature and timing actually serve the process.

Separate “available thermal output”, “transferred to storage”, “withdrawn from storage”, and “usefully consumed”. Where hot water recirculates through a tank, a flow meter may repeatedly count the same thermal energy unless the boundary is carefully defined. Heat exchanger losses, tank standby losses and mixing valves require appropriate treatment.

07. Core thermal calculations

For a single-phase liquid, instantaneous thermal power is Q̇ = ṁ × cp × (T_out − T_in), with mass flow in kg/s, specific heat in kJ/(kg·K), and Q̇ in kW. For water at ordinary process temperatures, an illustrative approximate volumetric expression is Q̇(kW) ≈ 1.16 × V̇(m³/h) × ΔT(K). This factor is only an approximation; use temperature-dependent density and heat capacity for the actual fluid, especially glycol mixtures.

Useful energy is the time integral of accepted sink thermal power: E_useful = Σ Q̇_useful,i × Δt_i in kWh_th, excluding bypass, unavailable demand and non-useful operating states. The comparison period, missing-data treatment, integration interval and clock alignment must be stated. Do not multiply a single peak thermal power reading by all compressor running hours.

A simplified avoided-fuel model is E_fuel,avoided ≈ E_useful / η_marginal where η_marginal is the applicable heating-system marginal useful efficiency under comparable conditions. It is a model, not measured fuel savings, unless the fuel boundary and counterfactual have been verified. A heat pump or electric heater requires its own performance basis and electricity accounting.

08. Measurement and uncertainty architecture

For Q̇ proportional to flow × temperature difference, a first-order relative uncertainty estimate for independent inputs is u_r(Q̇) ≈ sqrt[u_r(flow)² + (u(ΔT)/ΔT)² + u_r(cp)²]. Temperature-difference uncertainty can dominate at small ΔT. Sensor errors may be correlated, so a project-specific uncertainty budget must include calibration, installation, drift, time matching and covariance where material. No universal acceptance percentage is asserted.

09. Project timeline and before/after comparison

09.1 Baseline capture

Document existing compressor operating profile, heating fuel/electricity, hot-water demand, process temperature, ambient and production context. Establish whether historical heat demand can be matched to post-retrofit conditions.

09.2 Design and change control

Archive recovery exchanger specification, materials, pressure ratings, process isolation, relief/expansion arrangement, backflow prevention and the approved control sequence. Obtain OEM and qualified designer signoffs where needed.

09.3 Installation and safety testing

Retain pressure/leak tests, flushing, water quality checks, electrical isolation, insulation, functional interlocks and commissioning reports appropriate to the actual installation.

09.4 Initial functional demonstration

Measure source and sink flow/temperatures, compressor states, cooling protection, bypass transitions and backup heating response over representative conditions.

09.5 Comparable outcome period

Measure thermal delivery, process demand, auxiliary power and displaced heating input. Exclude unapproved or non-representative windows transparently.

09.6 Persistence check

Revisit exchanger fouling, flow control, seasonal use, changed process demand, control overrides, sensor drift and any maintenance impacts.

09.7 Claim and publication review

Reconcile actual data, uncertainty, ownership, consent, adverse outcomes, commercial relationships and final release status.

10. Mandatory asset — Heat Recovery Project Evidence Template

Every record field must contain an evidence ID or an explicit “not available”, verification method, result/status, owner, open item and disposition. A populated fictional example is not a real case. The template is the required project deliverable, not proof of completed work.

Controlled row states: NOT RECEIVED, RECEIVED, VERIFIED, EXCEPTION, REJECTED, NOT APPLICABLE. For each field add actual evidence reference, measurement method, result, responsible owner, open item and disposition. No real-world claim may be marked verified solely because a numerical worked example is supplied below.

11. Worked engineering case A — useful hot-water heat (illustrative)

Assume a fictional recovery loop delivers water at 3.6 m³/h with measured temperature rise of 18 K across the useful sink boundary. Using the illustrative water factor, useful instantaneous thermal power ≈ 1.16 × 3.6 × 18 = 75.17 kW_th. If this duty is accepted continuously for a verified 2.5 h interval, integrated useful energy is approximately 187.9 kWh_th. Neither number demonstrates actual fuel savings, annual performance or a completed Reddy project.

12. Worked case B — available heat versus accepted demand (illustrative)

Suppose a recovery exchanger can transfer 70 kW_th during a six-hour compressor operating period, but the process needs 50 kW_th for only four overlapping hours. Without useful storage, the maximum process-accepted energy is 50 × 4 = 200 kWh_th, not 70 × 6 = 420 kWh_th. Any claim of 420 kWh_th recovered for process use would overstate the evidence by 220 kWh_th. Real values must be derived from synchronised metering and the bypass log.

13. Worked case C — thermal storage accounting (illustrative)

A tank is charged with an illustrative 140 kWh_th, subsequently delivers 118 kWh_th to the process and loses or retains the remaining 22 kWh_th. If the claim boundary is useful process delivery, report 118 kWh_th. Do not count 140 plus 118 as 258 kWh_th. The actual storage state at the beginning and end of the interval must be reconciled to avoid shifting energy between periods.

14. Worked case D — fuel displacement and efficiency (illustrative)

For 300 kWh_th of verified useful heat, a model using an assumed marginal boiler efficiency of 0.82 estimates 300 / 0.82 = 365.9 kWh of fuel input displaced. This is a conditional estimate, not a fuel-meter reading. If the boiler cannot modulate, maintains standby losses or has changed return temperature, actual fuel displacement may differ substantially. Record boiler control state and measured fuel data where available.

15. Worked case E — auxiliary electricity and net benefit (illustrative)

A process accepts 210 kWh_th over an interval while recovery pumps and controls use 12 kWh_e. Gross useful heat remains 210 kWh_th; it is not correct to subtract electrical kWh from thermal kWh and label the result “198 kWh saved” without a defined conversion and counterfactual. Report 210 kWh_th delivered and 12 kWh_e auxiliary input separately, then calculate net cost or primary-energy effect using approved energy prices and factors.

16. Worked case F — uncertainty at small temperature rise (illustrative)

Suppose flow has illustrative standard relative uncertainty 2%, and each temperature sensor has independent standard uncertainty 0.2 K. For a measured ΔT of 3 K, u(ΔT) = sqrt(0.2² + 0.2²) = 0.283 K, or 9.43% relative. Ignoring cp uncertainty, combined relative uncertainty is approximately sqrt(2² + 9.43²) = 9.64%. At ΔT = 15 K with the same sensors, the temperature contribution is about 1.89%; instrument selection and installation therefore matter. Actual correlations require evaluation.

17. Worked case G — seasonality and annualisation (illustrative)

A measured winter week may show 900 kWh_th/day of useful heat, but summer process demand could be only 120 kWh_th/day. Multiplying the winter daily figure by 365 gives 328,500 kWh_th/year and would be indefensible without seasonal load evidence. Annual estimates require actual demand calendars, compressor availability, storage and bypass behaviour, and an explicit uncertainty range. Label any unmeasured future period as projected.

18. Worked case H — heat transfer with failed process acceptance (illustrative)

An exchanger transfers 60 kW_th into a water circuit, but process inlet temperature remains below an approved minimum because of a mixing valve fault. The exchanger’s thermal transfer can be measured, yet useful process heat may be zero for the affected interval if the sink cannot use it.

19. Worked case I — process throughput confounder (illustrative)

After a retrofit, boiler fuel consumption falls 11% while product output falls 18%. Without adjustment, the fuel decline may simply reflect lower demand. Compare equivalent throughput, product mix, water inlet temperature, heating duty, weather and boiler efficiency; use a defensible baseline model where matched periods are impossible. The heat recovery meter can establish delivered thermal energy independently, but avoided-fuel attribution remains bounded.

20. Worked case J — cooling penalty (illustrative)

A modified air-cooled package appears to supply 35 kW_th to a room-air duct, but increased backpressure and recirculation cause elevated compressor temperature and a 4 kW electrical fan penalty. The apparent thermal output cannot justify the installation without compressor OEM cooling acceptance, safe room ventilation and net energy analysis. Record the adverse result and any rollback; do not hide the penalty from the evidence case.

21. Confounder and causal-claim register

22. Risk, integrity and commissioning evidence

The actual engineering design must address exchanger pressure and temperature ratings, cross-contamination risk, approved fluid segregation, expansion and relief, scalding exposure, isolation and draining, water treatment, corrosion, fouling, condensate, freeze protection, legionella or hygiene controls where relevant, and safe fallback cooling. Applicability depends on the fluid, jurisdiction, process and OEM. Do not claim that this publication substitutes for design, inspection or statutory approval.

Commissioning should capture stable compressor cooling temperatures with recovery on and off; bypass operation at no sink demand; high-temperature and low-flow alarms; electrical failure; sensor fault; pump failure; and process-temperature acceptance. Record only the tests actually completed. Any unsupported protection or unresolved OEM limit is a release hold, not a small editorial issue.

23. Claim hierarchy and evidence classification

24. Customer ownership, consent and disclosure

Customer-provided values must not be treated as independently verified without reviewing the measurement method and source. Preserve withdrawal and correction procedures, including removal of structured data and indexed copies when appropriate.

25. Evidence visualisation requirements

Publish a boundary schematic and time-aligned traces of compressor input, recovery source and sink temperatures, sink flow, accepted thermal power, bypass status and process demand. Show missing intervals, calibration status, units, time zone and uncertainty. Use comparable axes for baseline and after plots. Provide text alternatives describing the actual measured conclusion, not marketing copy.

26. International standards and applicability plan

Candidate reference families for a project-specific review may include ISO 11011 for compressed-air energy assessment, ISO 50015 for energy-performance measurement and verification, ISO 5167 where differential-pressure flow measurement is actually applicable, and relevant national pressure-equipment, electrical and water-system requirements. Manufacturer installation and operating instructions remain controlling for compressor cooling integration. Do not state an edition, clause, certification or statutory compliance without the exact applicable source, jurisdiction, qualified review and retained record.

28. Expert FAQs

What makes a heat recovery project verified?

Source-backed, consented measurements of useful sink delivery with defined dates, boundaries, instrument uncertainty and process acceptance.

Can motor power be treated as recoverable heat?

No. Electrical input does not establish accessible heat at the required temperature, demand overlap or usable delivery.

Is heat transferred by the exchanger equal to useful process heat?

Not necessarily. Bypass, storage losses, return mixing and unavailable demand may reduce useful delivery.

How is recovered heat measured?

Measure suitable fluid flow and paired temperatures at the defined boundary, then integrate thermal power over accepted operating periods.

Why must flow and temperature measurements be synchronised?

A changing process can make unsynchronised readings produce an inaccurate heat-rate calculation.

Can a single peak kW_th value prove annual savings?

No. Annualisation needs compressor availability, sink demand, seasonal variation, bypass and storage evidence.

Does recovered heat automatically reduce fuel use?

No. Boiler control, marginal efficiency, standby and process changes determine actual displacement.

What is the role of compressor OEM approval?

To confirm that recovery integration preserves cooling, controls, pressure, temperature and warranty conditions for the exact asset.

How should auxiliary pump energy be shown?

Separately as electrical input, with gross thermal delivery and net economic or energy effects clearly distinguished.

What if a heat exchanger transfers heat but the process rejects it?

Report the exchanger transfer and failed useful-delivery acceptance; do not claim successful process heat recovery.

Can hypothetical calculations be published?

Yes, with prominent illustrative labels, assumptions and no implication of a real customer installation.

What blocks public release?

Missing source ownership, consent, instrument basis, claim review, process acceptance or unresolved engineering risks.