The OEM versus aftermarket argument is usually conducted as a loyalty question. It is really a specification question. A part either meets the drawing or it does not, and the brand on the box tells you nothing about the hardness of the running face inside it.
This article sets out what actually has to match, how to accept or reject a shipment on evidence, and where the commercial risk genuinely sits — so that the decision can be made on numbers rather than on habit.
Four categories are routinely confused:
The distinction that matters for a buyer is not the label but whether the supplier can produce a drawing, a material specification and a test result. Category two and three suppliers who can do this are safe partners. Category four suppliers who cannot should be avoided regardless of the name on the invoice.
Compatibility is not binary. Some features are functionally critical, others are cosmetic or logistical.
| Feature | Must match exactly | Tolerance / note |
|---|---|---|
| Bore and piston clearance | Yes | Micrometre-level; the single most important dimension |
| Valve plate kidney geometry | Yes | Timing, noise and flow depend on it |
| Face flatness and roughness | Yes | A rougher face will not bed and will leak |
| Material and hardness | Yes, within a stated range | Verify on incoming inspection |
| Splines, keys, flange pattern | Yes | Fit is not adjustable |
| Rotation direction and port angle | Yes | A left-hand part will not serve a right-hand machine |
| Surface finish of non-running faces | No | Cosmetic differences are acceptable |
| Packaging and markings | No | Relevant to logistics, not to function |
| Control regulator calibration | Yes, on assembled units | Settings must be verified, not assumed |
The last row is where many "compatible" replacements fail in practice. The mechanical part fits, the control is set to a different pressure, and the machine behaves as though the pump is worn.
Price comparison alone is misleading, because the costs are distributed unevenly in time. A simple framework makes the trade-off explicit.
| Cost element | Genuine OEM | Verified aftermarket | Unverified aftermarket |
|---|---|---|---|
| Purchase price | Highest | Moderate | Lowest |
| Lead time | Often longest | Usually short | Short |
| Incoming inspection burden | Low | Moderate, sampled | High, and often impossible |
| Risk of repeat failure | Low | Low if specified | High and unpredictable |
| Downtime exposure | Low | Low | High, multiplied by repeat events |
| Warranty recourse | Established process | Depends on terms | Often nominal |
For a machine earning revenue daily, one repeat failure usually erases the saving on several part sets. That is the honest case for buying verified parts — not brand loyalty, but the arithmetic of downtime.
Full inspection of every part is uneconomic. Sampling is not, provided it is structured and recorded.
| Level | Scope | When |
|---|---|---|
| Documentation | Certificate of material, hardness report, dimensional report | Every shipment |
| Visual | Packaging integrity, corrosion protection, part marking, obvious damage | Every shipment, 100 percent of cartons |
| Dimensional sample | Bore, piston, flatness, key features | One part per batch, more on a first order |
| Hardness sample | Portable tester on wear faces | One part per batch |
| Functional | Assembly into a test pump or a known-good machine | New supplier, new part number, or after any nonconformity |
Record the values, not the verdict. A sheet that reads "flatness 1.8 µm, hardness 60 HRC, bore 25.001 mm" allows a trend to be seen across shipments; one that reads "OK" does not.
Warranty terms are only as useful as the traceability behind them. Before ordering, establish that each part can be traced to a production batch, that the batch can be traced to material certificates, and that the supplier's remedy for a nonconforming part is replacement rather than negotiation.
Three clauses are worth insisting on. First, a defined claim window measured from delivery, not from installation, since installation may be delayed. Second, an agreed failure-analysis process, so that a claim is resolved by measurement rather than by assertion. Third, a statement of what voids the warranty — usually contamination, overpressure beyond the rated figure, or fitting without flushing — so both sides know the rules before an incident.
One further case deserves its own section because it produces failures that neither supplier will accept responsibility for. A repair shop fits a new aftermarket cylinder block with the original valve plate, or a new piston set into a used block. Each part may individually meet its own drawing, and the assembly still fails, because the block and plate are a bedded pair and the pistons in a set are lapped to a common height.
The rule is simple and worth writing into the workshop procedure: bedded pairs are replaced as pairs, pistons are replaced as sets, and a hybrid repair is only justified when both halves have been measured and fall inside the original tolerance band. Without measurement, a hybrid repair is a bet placed with the customer's machine as the stake.
There are clear cases where a verified aftermarket part is the better commercial decision: older machines whose OEM parts carry long lead times, fleets where a stock programme reduces downtime, and repairs where the OEM option is a complete assembly and only one component has failed. Kawasaki K3V pumps, for example, are widely supported by independent parts, and the K3V45 to K3V140 range is a routine repair for excavator fleets. Closed-circuit drives in Rexroth A22VG installations use A22VG45 and A20VG45 spare parts where availability often decides the repair route, and A28VO130 excavator pump spares cover a frame size where OEM lead times are frequently measured in months.
To make the comparison properly, send the pump model and nameplate data, the part numbers in question, the machine model, and the quantity you would hold as stock. We quote with the material and hardness specification stated, the traceability offered, and the warranty terms in writing — so the decision can be made against the OEM offer line by line.
It depends on the machine manufacturer's terms and the jurisdiction. Check before ordering, particularly on machines still inside the original warranty period.
Not on bedded pairs such as block and plate, and not with pistons from a different set. Mixing matched components is the most common cause of early failure in hybrid repairs.
Insist that each quotation states material grade, hardness range, flatness limit, whether pairs are matched, what documentation is supplied and the warranty terms. Compare those columns, then compare price.
Yes, if the rebuilder provides measurement records for the rotating group and a test report with flow, pressure and case drain at rated conditions. Without those documents, a refurbished pump is an unknown quantity.
The OEM versus aftermarket argument is usually conducted as a loyalty question. It is really a specification question. A part either meets the drawing or it does not, and the brand on the box tells you nothing about the hardness of the running face inside it.
This article sets out what actually has to match, how to accept or reject a shipment on evidence, and where the commercial risk genuinely sits — so that the decision can be made on numbers rather than on habit.
Four categories are routinely confused:
The distinction that matters for a buyer is not the label but whether the supplier can produce a drawing, a material specification and a test result. Category two and three suppliers who can do this are safe partners. Category four suppliers who cannot should be avoided regardless of the name on the invoice.
Compatibility is not binary. Some features are functionally critical, others are cosmetic or logistical.
| Feature | Must match exactly | Tolerance / note |
|---|---|---|
| Bore and piston clearance | Yes | Micrometre-level; the single most important dimension |
| Valve plate kidney geometry | Yes | Timing, noise and flow depend on it |
| Face flatness and roughness | Yes | A rougher face will not bed and will leak |
| Material and hardness | Yes, within a stated range | Verify on incoming inspection |
| Splines, keys, flange pattern | Yes | Fit is not adjustable |
| Rotation direction and port angle | Yes | A left-hand part will not serve a right-hand machine |
| Surface finish of non-running faces | No | Cosmetic differences are acceptable |
| Packaging and markings | No | Relevant to logistics, not to function |
| Control regulator calibration | Yes, on assembled units | Settings must be verified, not assumed |
The last row is where many "compatible" replacements fail in practice. The mechanical part fits, the control is set to a different pressure, and the machine behaves as though the pump is worn.
Price comparison alone is misleading, because the costs are distributed unevenly in time. A simple framework makes the trade-off explicit.
| Cost element | Genuine OEM | Verified aftermarket | Unverified aftermarket |
|---|---|---|---|
| Purchase price | Highest | Moderate | Lowest |
| Lead time | Often longest | Usually short | Short |
| Incoming inspection burden | Low | Moderate, sampled | High, and often impossible |
| Risk of repeat failure | Low | Low if specified | High and unpredictable |
| Downtime exposure | Low | Low | High, multiplied by repeat events |
| Warranty recourse | Established process | Depends on terms | Often nominal |
For a machine earning revenue daily, one repeat failure usually erases the saving on several part sets. That is the honest case for buying verified parts — not brand loyalty, but the arithmetic of downtime.
Full inspection of every part is uneconomic. Sampling is not, provided it is structured and recorded.
| Level | Scope | When |
|---|---|---|
| Documentation | Certificate of material, hardness report, dimensional report | Every shipment |
| Visual | Packaging integrity, corrosion protection, part marking, obvious damage | Every shipment, 100 percent of cartons |
| Dimensional sample | Bore, piston, flatness, key features | One part per batch, more on a first order |
| Hardness sample | Portable tester on wear faces | One part per batch |
| Functional | Assembly into a test pump or a known-good machine | New supplier, new part number, or after any nonconformity |
Record the values, not the verdict. A sheet that reads "flatness 1.8 µm, hardness 60 HRC, bore 25.001 mm" allows a trend to be seen across shipments; one that reads "OK" does not.
Warranty terms are only as useful as the traceability behind them. Before ordering, establish that each part can be traced to a production batch, that the batch can be traced to material certificates, and that the supplier's remedy for a nonconforming part is replacement rather than negotiation.
Three clauses are worth insisting on. First, a defined claim window measured from delivery, not from installation, since installation may be delayed. Second, an agreed failure-analysis process, so that a claim is resolved by measurement rather than by assertion. Third, a statement of what voids the warranty — usually contamination, overpressure beyond the rated figure, or fitting without flushing — so both sides know the rules before an incident.
One further case deserves its own section because it produces failures that neither supplier will accept responsibility for. A repair shop fits a new aftermarket cylinder block with the original valve plate, or a new piston set into a used block. Each part may individually meet its own drawing, and the assembly still fails, because the block and plate are a bedded pair and the pistons in a set are lapped to a common height.
The rule is simple and worth writing into the workshop procedure: bedded pairs are replaced as pairs, pistons are replaced as sets, and a hybrid repair is only justified when both halves have been measured and fall inside the original tolerance band. Without measurement, a hybrid repair is a bet placed with the customer's machine as the stake.
There are clear cases where a verified aftermarket part is the better commercial decision: older machines whose OEM parts carry long lead times, fleets where a stock programme reduces downtime, and repairs where the OEM option is a complete assembly and only one component has failed. Kawasaki K3V pumps, for example, are widely supported by independent parts, and the K3V45 to K3V140 range is a routine repair for excavator fleets. Closed-circuit drives in Rexroth A22VG installations use A22VG45 and A20VG45 spare parts where availability often decides the repair route, and A28VO130 excavator pump spares cover a frame size where OEM lead times are frequently measured in months.
To make the comparison properly, send the pump model and nameplate data, the part numbers in question, the machine model, and the quantity you would hold as stock. We quote with the material and hardness specification stated, the traceability offered, and the warranty terms in writing — so the decision can be made against the OEM offer line by line.
It depends on the machine manufacturer's terms and the jurisdiction. Check before ordering, particularly on machines still inside the original warranty period.
Not on bedded pairs such as block and plate, and not with pistons from a different set. Mixing matched components is the most common cause of early failure in hybrid repairs.
Insist that each quotation states material grade, hardness range, flatness limit, whether pairs are matched, what documentation is supplied and the warranty terms. Compare those columns, then compare price.
Yes, if the rebuilder provides measurement records for the rotating group and a test report with flow, pressure and case drain at rated conditions. Without those documents, a refurbished pump is an unknown quantity.