A tank dewatering trial is valuable only when the reader can see what was measured, under which conditions and against what baseline. A positive quotation or one successful drain cycle is not enough to establish a repeatable outcome.
This framework helps engineers and site teams collect a useful record. It is not a published customer case study and it makes no performance claim. Keep confidential site information controlled and obtain permission before using names, images or results publicly.
Build a credible tank dewatering trial record
Write the question before collecting data
State the problem the trial is intended to examine. It might concern repeatability, operator attendance, residual free water, product at the drain or the time required for a cycle. Avoid starting with a preferred conclusion such as a saving or a guaranteed reduction.
Define what counts as success, what would count as no change, and what evidence would make the trial inconclusive. Agree on the measurement locations, units, sample methods and responsibilities before comparing a new method with the baseline.
Name the tank, product family and time period in the controlled project record. A public version may anonymise them, but removing confidential details should not also remove the operating context needed to interpret the result.
Record a representative baseline
Document the existing method and enough normal cycles to show variation. Record drainage frequency, time, water removed, residual water where measured, operator involvement, downstream observations, maintenance events and disposal route.
Capture the conditions that influence the result: tank level or head, product, temperature, throughput, weather where relevant, settling time, transfer activity and any changes in the incoming liquid. Avoid selecting only the cleanest or most troublesome cycles.
Preserve the method behind each number. A volume estimated from a sight glass is not the same as a calibrated meter reading. State where a result comes from, who recorded it and whether the instrument or sample method has a known uncertainty.
Describe the actual duty and intervention
Record the tank type, liquid pair, specific gravity, viscosity and temperature range, solids, available head, expected flow, pipework and downstream destination. Include the relevant drawings and identify any uncertainty in those values.
For a valve trial, record the model, material and approved drawing only after they are confirmed. Describe installation and commissioning by reference to the approved project documents. Do not use a generic marketing diagram as proof of the configuration installed at the site.
List changes made during the trial, who approved them and when. If the tank, product, operating procedure or receiving route changed at the same time, the record should say so. Multiple simultaneous changes make it harder to attribute an outcome to one component.
- Trial question and agreed success measure
- Baseline period and the number of comparable cycles
- Tank, liquids, temperature, head and pipework
- Model and intervention details confirmed against approved records
- Measurement methods, instruments, units and uncertainties
Compare like with like and report all outcomes
Compare similar cycles or explain why they differ. A wet week should not be compared with a dry period without accounting for the change in water load. A higher tank level can affect flow, while a different product can change density and viscosity.
Report the full set of agreed outcomes, including results that did not improve. If drainage time falls but residual water rises, both facts matter. If product carry-over is measured only at the end of a cycle, say so rather than implying continuous monitoring.
Distinguish measured changes from calculated estimates. State sample size, date range, exclusions and confidence limits where available. Do not turn a short trial into annual savings or emissions claims without a reviewed calculation and stated assumptions.
Use the same start and stop definition for each cycle. If one method begins when an operator opens a valve and another includes setup, sampling or cleanup time, report those components separately before comparing duration.
Separate product evidence from the site safety case
A trial record is not authorisation to change a tank procedure. The site owner remains responsible for hazard assessment, operating procedures, isolation, waste handling and any permit or environmental requirements. Use the approved management of change and work control systems.
For flammable liquid tanks, HSE HSG176 describes specific drainage precautions for fixed bulk storage. Applicability depends on the site and material. The trial should record the safeguards used without presenting a product as a substitute for them.
If a trial produces an unexpected discharge, leak, unsafe roof condition or other abnormal event, stop under the site procedure and retain the event record. Do not omit the event from a public summary because it complicates the story.
Set confidentiality and publication permissions
Before anything goes public, obtain written approval for the customer name, site, photographs, project numbers, operating data and quotations. One permission does not automatically cover every detail or future reuse.
Ask the engineering owner to verify the model, materials, configuration, calculations and technical statements. Ask the customer owner to approve identification and confidential information. Preserve source records so the published summary can be traced back to the approved evidence.
If permission or reliable outcome data is missing, publish a general method or lesson without identifying the customer or implying a result. A transparent limitation is stronger than a fabricated case study.
Finally, keep the internal trial record and public article separate. The internal record may contain equipment tags, drawings and raw measurements that should not be published. The approved public summary should make clear which details were removed and whether that affects how broadly its lesson can be applied. This protects the customer while preserving enough context for an honest technical explanation.
Frequently asked questions
How many cycles are enough for a trial?
There is no universal number. Agree a period that covers expected operating variation and report the sample size, exclusions and uncertainty so readers can judge the evidence.
Can an anonymised trial still be useful?
Yes, if the duty, method, baseline, measurement and limitations are specific enough to explain what was learned without revealing confidential site details.
Can estimated operator savings be reported?
Only with a transparent method, supported inputs and review. Label estimates as estimates and do not convert them into a guaranteed annual saving.
Can Belfield publish a customer name from an internal reference?
No. Obtain the customer permission required for that name, site, image, quotation and result before publication. An internal reference is not publication consent.
Sources and further reading
The Belfield sizing and selection guide lists the operating information required for application review.
HSE HSG176 provides tank storage guidance for flammable liquids within its stated scope.
Next steps
For a technical review, share the duty and measured evidence with the Belfield application team. Obtain customer and engineering permission before publishing any named case. For more about PMA, visit the PMA website.
