Pump Reliability & Predictive Maintenance (PlusSmart) · 9 min read · Updated November 2026
Predictive Maintenance ROI: A Calculation Template with Real Numbers
Every predictive-maintenance proposal claims a return. Few show the arithmetic. This is the template we use with plants in Vietnam and across ASEAN — downtime cost, repair cost, energy, spares and labour on one side; sensors, platform and engineering on the other — with the figures from documented PlusSmart projects where they exist and conservative assumptions where they do not.
The case for condition monitoring is usually made with a single dramatic avoided failure. That is persuasive once and useless for a budget. A defensible ROI uses the plant's own failure history, prices every term, and separates what is certain (fewer repairs, less energy) from what is probable (avoided downtime). Below is the template; fill it with your numbers and it becomes the business case. Where PlusSmart has documented results we quote them; otherwise we use conservative industry ranges.
Step one: the baseline cost of the current state
| Term | How to compute | Example (20 critical pumps, mid-size plant) |
|---|---|---|
| Repair cost per year | Work orders: parts + labour + contractor + crane, for the fleet | 18 seal and bearing repairs × USD 2,500 = 45,000 |
| Downtime cost per year | Unplanned events × hours × cost of lost production (or overflow/penalty cost) | 6 events × 8 h × USD 1,500/h = 72,000 |
| Secondary damage | Failures that took other parts with them (motor, impeller, pipework) | 2 × USD 8,000 = 16,000 |
| Energy waste | kWh above a well-operated reference (specific energy benchmark), × tariff | 600 MWh × USD 0.09 = 54,000 |
| Spares held for emergencies | Carrying cost of inventory kept only because failures are unpredictable | USD 120,000 stock × 15 % = 18,000 |
| Total baseline | USD 205,000 / year |
Most plants discover during this step that their maintenance system does not record cause or downtime hours. A Pump Health Check from PlusSmart Analytics reconstructs the baseline from the CMMS, historian and shift logs; it is often the first time a plant sees the pump fleet as a cost.
Step two: expected reductions, by mechanism
| Mechanism | Conservative | Documented |
|---|---|---|
| Seal failures eliminated by dead-head/dry-run protection and flow balancing | −50 % | −100 % (iron-ore port: zero since 2015; AUD 350k → < 5k/year) |
| Unplanned stops converted to planned work by vibration/oil/ESA warning | −40 % | 52 → 0 per year (port); 5 of 9 assets flagged for action in first assessment (water plant) |
| Energy from correct operating point and fewer pumps running | −5 to −10 % | +30 % beyond target (24-pump geothermal plant) |
| Repair cost per event (planned vs emergency) | −30 % | Typical industry range −30 to −50 % |
| Emergency spares inventory | −20 % | Depends on lead-time policy |
Applying the conservative column to the example: repairs 45,000 → 22,500 (fewer failures) and the remaining events 30 % cheaper (→ ~15,750); downtime 72,000 → 43,200; secondary damage 16,000 → 8,000; energy 54,000 → 50,000; spares 18,000 → 14,400. New total about USD 131,000; annual saving ≈ USD 74,000 with conservative assumptions, three times that if the documented results are achieved.
Step three: the cost side
| Item | One-off | Annual |
|---|---|---|
| Wireless vibration sensors, 2 per pump × 20, gateways, installation | USD 20–30k | Platform and remote diagnosis USD 5–10k |
| Drive-based protection and balancing on one multi-pump station | USD 15–40k depending on scope | Minor |
| ESA on 10 inaccessible motors | USD 15–25k | Included in diagnosis service |
| Pump Health Check and engineering | USD 5–15k | Quarterly reviews USD 5–10k |
| Internal time (planner, technicians) | — | 0.2 FTE |
Indicative ranges; actual quotations come from PlusSmart Technology for the specific scope. For the example plant: one-off USD 60–100k, annual USD 15–25k, against a conservative saving of USD 74k/year — payback in the second year at worst, within the first year if the station control alone removes the seal failures. Pilot projects on 5–10 machines routinely show returns above 5:1 because they are aimed at the known bad actors.
What makes the number real
- Choose machines by consequence; a sensor on a pump nobody cares about returns nothing.
- Act on diagnoses; the saving comes from the alignment and the seal change, not from the alarm.
- Measure before and after — vibration, failures, kWh/m³ — and report quarterly; the second-year budget depends on it.
- Fix control first where VFDs exist; it is the cheapest term with the largest documented effect.
Send PlusSmart Vietnam your work-order export and we return the baseline and a filled template for your plant; the proposal that follows is sized to the pumps that carry the cost.
Need engineering help with this?
Our engineers apply this to real projects every week — selection, specification, commissioning. Send your question or a part number.
Quick answers
Our CMMS does not record downtime hours. Can we still build the case?
Yes. Shift logs, production records and operator interviews reconstruct events; PlusSmart Analytics accepts handwritten and irregular data. The first ROI uses estimates; the second year uses measured values.
Which saving is most certain?
Seal and bearing repairs avoided by fixing the operating condition (dead-head, dry running, balancing), because the mechanism is physical and immediate. Downtime savings depend on acting on warnings in time.
Is a small plant with 10 pumps worth it?
If one of them stops production, yes — a pilot on the three worst pumps with drive protection and sensors is inexpensive and the return is visible within a quarter.