How one Canadian producer found a cost variance its vendor reports had never shown, and what every operator can do about it.
Production chemicals are one of the largest controllable line items in lifting cost, yet most operators manage them with less rigor than they apply to a single compressor. Paraffin inhibitors, corrosion inhibitors, H2S scavengers, scale inhibitors, biocides and demulsifiers run continuously across thousands of injection points. The dosing target is a number in parts per million. The actual dose is whatever the pump delivered against whatever the well produced that day.
The gap between those two numbers is treatment variance, and in most programs nobody measures it independently.
Who is measuring whom
Ask a production engineer how the chemical program is performing and the answer usually arrives as a monthly PDF from the chemical vendor. It reports product delivered, sites serviced, and a treatment summary. It is produced by the party being measured.
That is not a criticism of vendors. The problem is structural. The vendor knows what was pumped. The producer knows what was produced, but often only after production accounting closes the month. Lab results sit in a third system. Telemetry, where it exists, sits in a fourth. PPM is a ratio of chemical to fluid, and when the numerator and denominator live in different places, updated on different schedules, the ratio is an estimate at best.
Across the programs we have audited, treatment variance of 15 to 20 percent is typical. Half of that is over-treatment: chemical purchased that the well did not need. The other half is under-treatment: the paraffin event, the corrosion failure, the sour gas excursion that shows up as a workover six months later.
What the producer found
A major Canadian oil and gas producer brought us in to reconcile chemical spend against actual treatment across its program. The operator had multiple vendors, thousands of injection points, and a reporting process that looked healthy from the outside.
The first step was mechanical: pull every injection rate, every production allocation, every lab result and every telemetry feed into a single record, then compute target-versus-actual PPM per well, per vendor, per month.
The initial result was a PPM variance above 300 against target. Not 300 percent — 300 PPM of drift on wells that were supposed to be treated at a fraction of that. Some sites were running far over target. Others were effectively untreated while being invoiced as treated.
Once the measurement existed, the correction was fast. Vendors adjusted rates against a shared, independent number instead of their own reports. Production allocation moved from quarterly to monthly so PPM reflected current flow. Within the program, PPM variance fell from over 300 to 29, and the operator eliminated roughly $660,000 a month in chemical cost variance.
That money had always been in the budget. It had simply never been visible.
A second case, same pattern
A different producer running seven chemical vendors across its assets measured 15 to 20 percent treatment variance when we started. After twelve months on a unified target-versus-actual record, variance across all seven vendors settled at 8 percent. On its paraffin program specifically, where the operator was balancing inhibitor dosing against physical dewax events, combined treatment and dewax cost dropped by $0.78 per BOE and variance fell from 18 percent to under 5.
Two producers, different basins, different vendors. The same blind spot and the same fix.
Why hardware alone does not close the gap
The usual response to this problem is to buy more monitoring: tank-level sensors, smart pumps, injection-rate telemetry. Hardware helps, and it is worth having. But hardware measures the pump. It does not know what the well produced, what the lab found in the water sample, or what the target PPM should be this month.
The correction in both cases above came from the software layer, not the sensor layer. Specifically, from a chemical program management software platform that treats field capture, production allocation, lab analysis and telemetry as one dataset with one question: was every injection point treated to target, and what did it cost?
That layer has to be hardware-agnostic. Operators run a mix of pumps, telemetry vendors and SCADA systems, and most vendor-supplied monitoring only optimizes the devices that vendor sold. The producer’s record needs to sit above all of it.
Five things any operator can do this quarter
None of these require a software purchase. They require a decision to measure.
1. Write down the target PPM for every injection point. Not per well type or per field: per point. If the target does not exist as a number, variance cannot exist as a number either.
2. Allocate production monthly, not quarterly. PPM is chemical divided by fluid. A quarter-old denominator makes the ratio meaningless for operational decisions.
3. Log actual pump rates, not planned ones. A route sheet that says “set to 2 L/day” is a plan. The number that matters is what the pump delivered, captured on site or by telemetry.
4. Compare target to actual per well, per vendor, every month. This is the report that vendor PDFs do not contain. It is also the only report that tells you where the money is going.
5. Benchmark vendors on treatment accuracy, not unit price. A vendor at $4 a litre who hits target beats a vendor at $3.50 who drifts 20 percent. Most procurement teams still buy on the second number.
The broader point
Energy companies have spent a decade instrumenting the physical asset. Production chemistry is one of the last large cost centres still run on vendor reports and spreadsheets, and the evidence from the field is that the variance hiding inside it is material: six figures a month for a mid-sized producer, with asset-integrity risk on the under-treated side that never shows up on the invoice.
The fix is not exotic. It is an independent record of target versus actual, owned by the producer, that every vendor is measured against. Once that record exists, the cost correction tends to take care of itself.
Michael Jones is CEO of Embark Innovations Inc., a Calgary-based company that builds Prodesy, a production chemical program management platform used by oil and gas producers and chemical vendors across Canada and the United States. He has worked in oilfield data and mobile field technology since 2004.