Why Gas Consumption Data Can Mislead a Gas-Engine Investment Decision

Key takeaways

• Gas volume is not energy until reference conditions and gas composition are known.
• Fuel and electrical measurements must cover the same time period and system boundary before efficiency is calculated.
• The EV-014 cumulative counters yield 20,938 engine-hours over 302 complete intervals between 303 reporting dates; the available electrical-production records total approximately 18,893 MWh on a separate record-coverage basis.
• The engine-hour total is calculated from cumulative column E. The electrical-production total remains a separate evidence item and is not used to claim actual efficiency.

Why can measured gas consumption still mislead an investment model?

A meter can record a real volume while the model applies the wrong pressure, temperature, composition, time interval or plant boundary. The review must distinguish measured volume, converted energy, benchmark assumptions and calculated efficiency, and must keep missing calibration or composition evidence visible.

What is the minimum evidence set for a fuel-efficiency claim?

Use timestamped gas and electrical records for the same interval, stated reference conditions, gas composition or conservative heating-value basis, meter identity and boundary, calibration status, treatment and auxiliary loads, excluded periods and the calculation formula. Without that alignment, the result is a screening estimate rather than a verified efficiency.

Gas-consumption data can look precise and still be unsuitable for an investment decision. A totalizer value, spreadsheet column or monthly invoice does not automatically show what passed through the meter, at which reference conditions, with which gas composition, or during the same time window as the electrical output.

K4-Technology encountered this issue while analysing 303 reporting dates from a three-unit J320 installation. Gas consumption was recorded by site meters, but the implied consumption was materially below the passport reference. K4-Technology did not use the meter-derived figure to claim actual efficiency; the investment calculation used passport consumption instead. The exact meter value remains unpublished.

The commercial and technical problem

Fuel is often one of the largest variable inputs in a gas-to-power model. Small errors in assumed consumption or gas energy content can materially change operating-cost, margin and debt-service projections. The risk grows when a reported gas volume is treated as directly comparable across meters, periods or sites without checking its basis.

Why gas data becomes misleading

Volume is not energy

A volume in normal cubic metres becomes an energy input only after the applicable gas composition or calorific value and reference basis are known. The same recorded volume can represent different energy content.

Reference conditions may differ

Metering systems and source documents may use different temperature, pressure or standard-condition conventions. Combining them without reconciliation creates a systematic error that can be difficult to see in a monthly total.

Time alignment can fail

Electrical energy, engine runtime and gas totals must cover the same boundary and interval. A site meter may include boilers, flares, start gas, other engines or periods not represented in the electrical dataset.

Meter quality and data handling matter

Calibration, rangeability, pressure/temperature compensation, resets, missing intervals, manual transcription and unit conversion can all affect the result. A plausible-looking number is not evidence that these issues were controlled.

Plant boundaries may not match

An engine performance curve describes a defined engine and test basis. A site meter can include treatment, auxiliary users or multiple units. Comparing the two without a boundary map is not a valid efficiency calculation.

What should actually be checked

• Meter tag, location and physical measurement boundary.
• Reported unit and reference conditions.
• Calibration and compensation status for pressure and temperature.
• Gas composition or calorific-value source and sampling frequency.
• Timestamp, interval and time-zone alignment with electrical production.
• Whether the gas total includes other consumers, flaring, venting or start-up periods.
• Data gaps, resets, negative values, duplicates and manual corrections.
• Consistency at daily, monthly and campaign level before an aggregate figure is accepted.

A verified K4-Technology example: useful data without a false efficiency claim

The reviewed dataset covered three J320 units over 303 reporting dates. Across the 302 complete intervals between those dates, the cumulative counters in column E yield 20,938 recorded engine-hours. The available electrical-production records total approximately 18,893 MWh on a separate record-coverage basis. These figures support runtime, generation and event-classification analysis without implying that every production entry is coterminous with the complete-interval set.

The meter records indicated consumption below the passport reference, so K4-Technology did not publish an actual efficiency claim from those values. For the investment calculation, K4-Technology used passport consumption rather than the lower meter-derived figure. This preserves a defensible evidence boundary.

How a reference benchmark can be used safely

For context only, the published case used three public reference points for a J320-type performance curve: 274 Nm³/h at 1,063 kW, 211 Nm³/h at 796 kW and 148 Nm³/h at 529 kW, with a stated lower heating value of 9.5 kWh/Nm³. Linear interpolation around 900 kW gives approximately 236 Nm³/h, or about 0.262 Nm³/kWh.

Applied mechanically to 18.893 GWh, that intensity would imply roughly 4.95 million Nm³. It is a benchmark estimate, not measured site consumption and not proof of actual efficiency. Its purpose is to test order of magnitude and expose reconciliation questions. Project decisions require the real gas boundary, composition and meter basis.

Common mistakes and red flags

• Dividing a site gas total by engine production without mapping the meter boundary.
• Using a nameplate or curve value as actual site performance.
• Mixing higher and lower heating values.
• Ignoring pressure, temperature and normalisation conventions.
• Using monthly totals to hide missing or contradictory daily data.
• Treating unexplained differences as rounding.
• Publishing an efficiency percentage because the spreadsheet produces one.

How K4-Technology approaches the problem

K4-Technology starts with an evidence map: measurement point, unit, reference conditions, timestamps, gas quality, electrical boundary and data owner. The analysis then checks completeness, reconciles intervals and separates measured values from engineering benchmarks. If the data cannot support an actual efficiency result, the report states the limitation and identifies the evidence needed to close it.

Practical decision checklist

• Can the gas meter and its boundary be identified?
• Are reference conditions and calorific value documented?
• Do gas and electrical records cover the same intervals?
• Are other gas users excluded or quantified?
• Does the data pass daily as well as aggregate checks?
• Are benchmark calculations labelled as benchmarks?
• Has uncertainty been carried into the financial model?

When remote review is sufficient

Remote review can identify boundary errors, unit problems, missing intervals, inconsistent totals and unjustified efficiency claims when meter records, electrical data and supporting documents are available.

When physical inspection is required

A site visit is required when meter identity, piping boundary, instrumentation, calibration status or other physical conditions cannot be established from reliable records. Sampling may also be necessary when gas composition is uncertain.

Conclusion

The best investment analysis does not force every supplied number into the model. It distinguishes measured facts, calculated results, reference benchmarks and unresolved gaps. That discipline prevents weak gas data from becoming a confident but misleading project return.

FAQ

Is a gas-meter reading enough to calculate efficiency?

No. The reading needs reference conditions, heating value or composition, time alignment, meter identity, calibration context and a matching electrical-energy boundary. Otherwise the calculation may compare different quantities.

Why use passport figures when measured consumption is lower?

A conservative screening model may use the minimum supported passport value when lower field consumption has not been reconciled to reference conditions, gas composition, meter accuracy and system boundary. The measured result should remain a separate evidence item.

What basis is used for the operating example?

The article uses the EV-014 cumulative-counter basis of 20,938 engine-hours over 302 complete intervals between 303 reporting dates. The available electrical-production records total approximately 18,893 MWh on a separate record-coverage basis.

Which model-review service applies?

An Express Model Review starts at USD 450 and a Full Model Review starts at USD 1,500. The appropriate scope depends on the model, evidence set and decision.