3. Bid evaluation — TCO framework and scoring matrix
Pumps lose 70-90% of their lifecycle cost to energy and downtime during failures, not to capital. A bid evaluation that scores capital heavily and energy lightly will produce the cheapest pump up front and the most expensive pump over its life.
3.1 The TCO framework
Total Cost of Ownership over the pump’s design life:
TCO = CAPEX + ENERGY + MAINTENANCE + DOWNTIME + DISPOSAL
Typical breakdown for a 50 kW industrial centrifugal running 8.000 h/yr over 15 years in Brazil (R$ 0.65/kWh industrial tariff, 2026):
| Component | Calculation | R$ (typical) | % of TCO |
|---|---|---|---|
| CAPEX | Purchase + freight + install | 250.000 | 8% |
| Energy at 75% pump efficiency | 50/0.75 × 8.000 × 15 × 0.65 | 5.200.000 | 73% |
| Maintenance (annual ≈ 4% CAPEX) | 250.000 × 0.04 × 15 | 150.000 | 2% |
| Spare parts (2 sets impeller + seal over 15 yr) | _ | 80.000 | 1% |
| Unplanned downtime (1 event × 5 days × 4.000 R$/h) | _ | 480.000 | 7% |
| Decommissioning | _ | 30.000 | < 1% |
| Risk-adjusted reserve (process-criticality × 5%) | _ | 600.000 | 8% |
| Total | 6.790.000 | 100% |
The headline: a R$ 50.000 difference in CAPEX is dwarfed by a 5-percentage- point efficiency difference (which moves energy by R$ 350.000 over 15 years).
3.2 Scoring matrix template
Build a weighted scoring matrix. Recommended weights for industrial-process duty:
| Criterion | Weight | Why |
|---|---|---|
| Lifecycle energy cost | 30-40% | Dominant component of TCO |
| Spare-parts lead time + local stock | 15-25% | Hedge against unplanned outage |
| CAPEX | 15-25% | Real money but not the biggest |
| References at duty + industry | 10-15% | Reduces hidden-failure risk |
| Documentation completeness | 5-10% | Acceptance gating |
| Warranty terms | 5-10% | Insurance against year-one failure |
| Compliance / certifications | 5-10% | Legal / regulatory required |
For non-process duty (utility, fire, ancillary), shift weights:
- Fire pumps: compliance certifications dominate — UL / FM listing + INMETRO chain are weighted 30%+; energy is irrelevant (operates rarely)
- Standby / spare: CAPEX is most relevant; energy weight drops to 5-10%
- Critical-process (no redundancy): spare-parts and references rise; both should be 25%+
3.3 How to score energy fairly
Vendors quote efficiency at rated point. To compare on lifecycle cost, you need efficiency at the actual operating envelope, not just at rated:
Annual energy = sum over operating conditions of:
(Q × H × ρ × g) / (η × 3.6e6) × hours_per_year
For a vendor’s bid, request:
- Efficiency at minimum, normal, rated, and maximum operating points
- Single weighted-average efficiency (vendor’s best estimate at duty cycle)
- Pump curve plot showing efficiency line, not just head line
A vendor that publishes only rated efficiency is hiding off-design performance. The off-design performance is where most pumps actually run.
3.4 Lead time scoring
Two numbers per bid:
- Initial delivery lead time — order to factory acceptance
- Spare-parts replenishment lead time — typical order to delivery for a critical wear part (impeller, seal, bearing)
Local manufacturers typically quote 8-12 weeks initial / 4-8 weeks parts. Major international manufacturers with Brazilian assembly: 12-20 weeks initial / 8-16 weeks parts. Pure-import: 16-24 weeks initial / 16-24 weeks parts.
For a critical-service pump without redundancy, multiply parts lead time by the average failure-event cost per day to get the embedded annual risk.
3.5 References scoring
Score on three sub-criteria:
| Sub-criterion | Weight | How to score |
|---|---|---|
| Reference at similar duty point (±30% Q, ±20% H) | 40% | 0 / 50 / 100 (none / single / multiple) |
| Reference in same industry vertical | 30% | 0 / 50 / 100 (none / partial / direct) |
| Reference contact reachable + cooperative | 30% | 0 / 50 / 100 (refused / handled / direct) |
A vendor with three direct industry-and-duty references that you can call yourself scores 100/100 here. A vendor with a glossy brochure of unnamed “Fortune 500 clients” scores 0/100.
3.6 Compliance scoring
Binary on each required certification — vendor either has it for the specific model proposed, or does not:
- API 610 / API 676 — if process duty, mandatory
- NFPA 20 / NBR 16704 — if fire, mandatory
- HI 14.6 Grade 1U — if API duty, mandatory; otherwise score on grade
- CRCC Petrobras — if Petrobras / oil-and-gas, mandatory
- INMETRO-recognized lab certificates — for Brazilian regulatory acceptance, mandatory
- CE marking — if exporting to EU end-user, required
A vendor missing any mandatory certification disqualifies for that service class. There is no partial credit.
3.7 Scoring matrix template
See templates/evaluation-matrix.md for
a ready-to-edit weighted matrix.
Worked TCO comparison example
Two bids for an 80 m³/h water-circulation duty, 8.000 h/yr, 15-year life:
| Bid A (low-CAPEX importer) | Bid B (local manufacturer) | |
|---|---|---|
| CAPEX | R$ 180.000 | R$ 235.000 |
| Efficiency at duty | 72% | 78% |
| Initial lead time | 22 weeks | 10 weeks |
| Spare-parts lead time | 18 weeks | 5 weeks |
| References at this duty | 1 | 4 |
| Documentation language | English (translation cost ~ R$ 8.000) | Portuguese |
Energy cost over 15 years (50 kW shaft assumption):
- Bid A: 50/0.72 × 8.000 × 15 × 0.65 = R$ 5.417.000
- Bid B: 50/0.78 × 8.000 × 15 × 0.65 = R$ 5.000.000
Energy savings on Bid B: R$ 417.000 over 15 years — already 7.5× the CAPEX delta.
Plus Bid B saves:
- 12 weeks earlier commissioning → R$ 100.000+ deferred-revenue gain
- 13 weeks faster spare-parts replenishment → reduces unplanned outage exposure
- Documentation translation cost not incurred → R$ 8.000
Total advantage of Bid B: R$ 470.000+ over 15 years. CAPEX delta of R$ 55.000 is the wrong frame to evaluate the bid.
This is not unusual. It is the typical result when TCO is computed honestly.
Next section: Contract terms — clauses that protect the buyer.