# Hospitality estate gas investigation

## Evidence pack

This pack records the evidence-led investigation in a form that can be reviewed by operational, finance, and energy teams. It keeps the useful findings and decision trail while excluding identifying details.

---

## Executive finding

The gas-use profile contained a persistent out-of-hours demand — a **ghost load** — that did not fall away as cleanly as the operating pattern suggested it should. The investigation therefore moved the question from “is the bill simply high?” to “what is still drawing gas when normal activity has reduced, and what evidence is needed before changing anything?”

The review supported an operational investigation and monitoring step. It did **not** support naming a single appliance, declaring a plant fault, or committing to replacement technology without first confirming the source of the overnight demand.

## 1. The question being tested

The starting concern was an unclear gas-use pattern. A high or uneven bill alone could not distinguish between:

- normal demand associated with the way the building was being used;
- a recurring base load outside the main operating period;
- a change in performance that needed operational context; or
- a fault, control issue, or schedule that still needed to be checked.

The investigation was designed to separate what the gas data actually showed from assumptions about why it was happening.

## 2. Data examined

The review brought the available consumption and operating evidence into one readable view. The examination covered:

1. The gas-consumption profile across the available record.
2. The relationship between daytime demand and overnight demand.
3. The shape of the normal daily profile, including the expected fall in use outside the main operating period.
4. Exceptions and changes that did not match the prevailing pattern.
5. Operational context needed to distinguish expected demand from avoidable demand.
6. The evidence required to test a cause before selecting an intervention.

The method was deliberately suitable for a spreadsheet or meter-data review first; a dashboard was not treated as a prerequisite for finding the pattern.

## 3. What the gas-use evidence showed

### A recurring overnight base load

The most important observation was continued gas use outside the main operating period. Instead of treating that demand as background noise, the review treated it as a measurable base load to be explained.

That pattern is the investigation’s concrete finding: there was a gas-use signal worth tracing, not enough evidence to jump straight to a particular fix.

### A useful comparison, not a single bill total

Comparing day and night use made the pattern easier to see than a period total on its own. The review focused on the shape and repeatability of the demand, because a recurring base load can point to a controllable operating condition even when the total bill looks ordinary or is affected by other factors.

### Exceptions mattered

The normal profile was used as a baseline. Periods that departed from it were treated as questions to investigate, not automatically as savings or faults. This protected the review from mistaking one unusual reading for a stable finding.

## 4. What was ruled in and ruled out

### Supported by the evidence

- A recurring out-of-hours gas load was present in the reviewed profile.
- The load was significant enough to justify tracing what remained in operation.
- Day-versus-night comparison and an average profile were useful ways to isolate the issue.
- Operational context was necessary before deciding whether the demand was expected or avoidable.
- The next decision should be based on confirming the source and response, then checking that the profile changes.

### Not established by the evidence

- A single named appliance or item of plant as the cause.
- A confirmed equipment failure.
- A case for immediate replacement, renewable generation, or other capital work.
- A reliable saving figure without first measuring the load, correcting the cause, and verifying the result.

This distinction is important. The investigation identified a worthwhile lead and a proportionate next action; it did not turn an unexplained pattern into an unsupported diagnosis.

## 5. Investigation method

The investigation followed a simple evidence-to-action sequence:

### Establish the baseline

Review the available gas data and establish the normal daily profile. Identify when demand rises, when it should reduce, and what a repeatable pattern looks like.

### Compare day and night

Compare daytime demand with the out-of-hours base load. A high overnight share is a prompt to look for energy left running, not by itself proof of a fault.

### Check exceptions against operations

Match unusual changes against operating conditions, occupancy, weather where relevant, maintenance, and known schedules. Keep the explanation open until the data and operating context agree.

### Trace before changing

Check controls, schedules, plant status, and the relevant operating process to identify what is creating the overnight demand. Record the finding so that a correction can be repeated and verified.

### Monitor the response

After the operational change, compare the new profile with the baseline. A reduction that persists is evidence of improvement; a change that does not persist is a prompt to continue investigating.

## 6. Recommended action

The recommended action was to **trace and confirm the overnight gas load before spending money on a larger intervention**.

In practical terms:

1. Confirm the periods in which the gas load continues outside normal operation.
2. Walk the relevant controls, schedules, plant status, and operating routines with the people who know how the building runs.
3. Identify what can safely be switched off, reset, or scheduled without affecting required operation.
4. Make the lowest-risk operational correction first.
5. Continue monitoring and compare the resulting profile with the established baseline.
6. Only then decide whether further engineering work or capital investment is justified.

This sequence turns the ghost load into a testable action. It also keeps efficiency work ahead of technology selection: understand and reduce avoidable demand before sizing or funding a new solution.

## 7. Decision record

| Decision point | Evidence-led position |
| --- | --- |
| Is there a pattern worth investigating? | Yes. The recurring out-of-hours load is a clear lead. |
| Is the cause proven? | No. The profile identifies the condition, not the individual cause. |
| Is immediate replacement justified? | No. Confirm the source and response first. |
| What should happen next? | Trace the load, make a proportionate operational correction, and monitor the result. |
| What would justify a larger project? | A confirmed cause, a measured avoidable load, and a verified case for the proposed intervention. |

## Closing note

The value of this investigation was not a speculative fix. It was a clearer demand picture, a defensible evidence trail, and a practical next step: find what is still using gas out of hours, test the lowest-risk response, and use the resulting data to decide what deserves further action.