Energy intensity denominator workbook
Before dividing energy by output, establish what the output number counts. A dispatch total may be complete and still be the wrong denominator for a production-based measure. This workbook preserves the raw quantity, records supported adjustments, compares the result with a separately compiled control and keeps the human alignment decision beside the ratio.
The editable XLSX contains a blank Denominator bridge, a separate Fictional example and a detailed Read me. It supports up to twelve adjustments for one numerator-denominator pair. The matching PDF is a static sheet reference and does not recalculate. All organisations, records, quantities and review references in the case are invented.
Download the energy intensity denominator workbook (XLSX).
Download the static energy intensity denominator workbook (PDF).
Name the measure before selecting an export
Write the denominator definition in words: what is counted, at what stage, for which population and over which period. Completed product, dispatched product, sales, occupied time and available capacity can all appear in operational records. They do not become interchangeable because their numbers are convenient or because an export happens to be available at the reporting deadline.
Record the energy definition and unit separately. Specify whether the quantity is electricity only or a wider energy total and identify its measurement and conversion basis in the retained schedule. This workbook accepts one prepared numerator; it does not add fuels, convert bills into energy or decide whether site or primary energy is appropriate. Those decisions must be resolved in the underlying method.
The U.S. Department of Energy distinguishes energy intensity from energy efficiency and discusses other influences on intensity. Its Energy Efficiency vs. Energy Intensity page provides context for keeping a simple ratio’s interpretation bounded.
Keep three denominator positions visible
Enter the raw denominator with its source version and precise locator. Then enter the signed adjustments needed to reach the selected definition. Finally, provide an independently compiled denominator control and its own source reference. Do not populate the control by copying the bridge result. A copy would make the comparison agree without adding a second evidence route.
Each adjustment needs an ID, reason, signed amount, source locator, counting rule and coverage note, owner, review state and review reference or next action. A negative amount removes a quantity from the raw basis; a positive amount adds one. The workbook does not select the amount or its sign. Keep the source record unchanged and make the transformation visible in the bridge.
State the expected adjustment count and retain the complete adjustment-list reference. Zero adjustments are supported, but enter zero deliberately and explain why the raw figure already matches the intended definition. A correct count alone does not establish that every adjustment has been identified. The external list and the independently compiled control remain important even when the sheet has no missing fields.
Separate numerical completeness from the decision
The numerical bridge becomes available when setup, row checks and expected count agree. It adds the signed adjustments to the raw denominator and displays the difference from the independent control. The running results beside individual adjustments are provisional. They help trace the calculation, but an intermediate value is not the completed denominator to carry into a report.
The ratio has further gates. The completed denominator must be nonnegative, agree exactly with the control and be positive before division. Alignment must be Confirmed, and every adjustment review must be Confirmed. If review is Pending, the numerical bridge can stay visible while the ratio is withheld. These states record supplied human decisions; the workbook does not verify the evidence or grant approval.
The alignment record should explain why numerator and denominator cover the intended boundary, period, activities and measurement basis. A full-plant energy figure divided by one product line’s output creates a different measure from the stated full-plant ratio. Likewise, a current-period energy figure cannot silently borrow a full-year denominator. Periods here are text labels, so those relationships need a human check.
Reconcile a fictional dispatch export to completed output
Alder Plant A’s fictional request is whole-plant energy per tonne of one defined finished product completed during June 2026. EN1 supplies 3,300 GJ for that plant and month. DS1 records 1,000 net tonnes dispatched during June. ST0 records 200 tonnes of opening finished-goods stock and ST1 records 300 tonnes of closing stock. PR1 separately records 1,100 tonnes completed during June.
The case stipulates the same product and mass basis across all stock and movement records. There are no returns, inter-site transfers, losses, rework or other finished-goods movements. The stock relationship is opening plus completed output, less dispatch, equals closing. These are simplifying teaching assumptions. A real stock account must identify any additional movements and cannot borrow this two-adjustment bridge without checking them.
Start with DS1’s 1,000 tonnes. Adjustment A01 subtracts ST0’s 200 opening tonnes, and A02 adds ST1’s 300 closing tonnes. The bridge is 1,000 − 200 + 300 = 1,100 tonnes completed, matching PR1. The cross-check is 200 + 1,100 − 1,000 = 300 tonnes closing stock. The net adjustment is +100 tonnes; both constituent adjustments remain visible.
The selected ratio is 3,300 GJ ÷ 1,100 completed tonnes = 3 GJ per completed tonne. Dividing by the unadjusted dispatch figure gives 3,300 ÷ 1,000 = 3.3 GJ per dispatched tonne. That is a differently labelled question. The energy quantity has not fallen, and the calculation supplies no intervention, efficiency improvement or product-specific energy allocation.
Keep the two stock endpoints rather than recording only the net +100. Another reader should be able to reconstruct why each endpoint has its sign and locate both source records. The bridge does not establish which individual dispatched items came from opening stock. It relies on the complete stock relationship under the stated assumptions, not an unsupported claim about the identity of particular units shipped.
Test the evidence that releases the ratio
Change A02’s review state to Pending in a copy of the example. The 1,100-tonne numerical bridge remains available, but the ratio disappears. Keep the pending action and owner with the closing-stock evidence. Setting the state back to Confirmed should reflect an actual recorded decision in a live file; a spreadsheet status is not the review itself.
Next, change the independent control to 1,090 tonnes. The bridge still gives 1,100 and shows a ten-tonne disagreement. The ratio is withheld. Ask the owners to compare product scope, timing, quantity basis and record versions. Do not change an adjustment to ten tonnes less simply because that would make the control check pass.
If an adjustment is deleted and the expected count is reduced, the independent control still tests the resulting denominator. In the example, removing the opening-stock subtraction leaves 1,300 rather than PR1’s 1,100. However, a preparer who also overwrites PR1 with the bridge result can defeat that protection. Preserve the separately compiled control and its provenance; the workbook cannot enforce independence outside the file.
Keep zero, missing and negative results distinct
A missing required quantity prevents the bridge from being presented as complete. A supported numeric zero remains a value. If a fully documented no-adjustment state has raw denominator and control both zero, the bridge is zero but the ratio is undefined. This remains true when the energy numerator is also zero. The tool never turns zero divided by zero into a zero intensity.
A negative bridge result is retained as a numerical issue to investigate and is never used as an intensity denominator. Keep the absolute energy quantity visible while resolving the denominator. If the chosen population includes an idle operation, do not remove its energy merely to obtain a convenient output ratio; apply the intended boundary and definition consistently in the supporting method.
Preserve the workbook’s input contract
Numeric fields accept literal numbers with at most three decimal places. Energy, raw denominator and control must be between zero and 1,000,000,000; each signed adjustment must be between −1,000,000,000 and +1,000,000,000. Derived denominator amounts are rounded to three places. Division uses the completed denominator without rounding the ratio, and General display may use scientific notation for very small values.
Keep General input formats and paste values only. Numeric formulas, logical values, text numbers, errors and excess precision are rejected. Date-formatted serials are still numbers, and these checks cannot detect a mistaken date or wrong semantic unit in a quantity field. There are no automatic conversions, calendar checks, macros or external data connections. Rows 19–30 are the entire adjustment capacity; extending them requires redesign and retesting.
Cache-cleared headless LibreOffice tests cover the exact candidate and its mutations. Microsoft Excel and desktop interaction have not been tested. Verify a saved copy in the application you intend to use, preserve a clean master and retain the numerical controls with the exported ratio. A static PDF cannot replace the recalculating working file.
Use the result with its full label
The intensity guide explains how to combine compatible quantities before division. The energy-comparison guide addresses the separate evidence needed for a saving claim. Use the calculation change register if the denominator correction affects an earlier output or a released statement. This workbook supplies the narrower raw-to-selected denominator record and independent-control check.
For a scoped reporting-record review, contact Nitium Advisory with the intended ratio, period and a non-confidential outline of the unresolved evidence.
Version 1.0. Prepared 15 October 2026. Original Nitium Advisory workbook, diagram and fictional case. Source context checked on that date. This optional working aid supplies no reporting-framework compliance conclusion, professional sign-off, verified saving, source endorsement or independent assurance.
Comments