I’ve pulled hundreds of failed seals out of pumps, and here’s what I’ve learned: the technician who can name every face material by color but skips the caliper measurements is the one who orders the wrong replacement. An Oceaneering reliability study across 40 North Sea pumps found related problems — variance in seal installation methods and low confidence in spare seal stock were among the root causes of repeated failures, and the fixes included installation training and a critical spares review.
Part numbers are a shortcut when they’re legible — and on most failed seals, they’re not. Color gets you in the ballpark on materials, but materials alone won’t tell your supplier what to ship. Configuration, dimensions, then materials — in that order. That’s the process that actually gets you the right replacement seal.
Identify the Configuration First
Before you measure anything, determine what category of seal you’re holding. A cartridge seal arrives as a complete, self-contained unit mounted on a sleeve — you’ll see set screws, a gland plate, and shipping tabs or spacers all pre-assembled. A component seal comes as separate pieces: a rotary unit, a stationary seat, and hardware you assemble in the field.
I’ve seen a technician install a cartridge seal without removing the shipping spacer, and the seal destroyed itself on startup. If you don’t recognize a cartridge seal for what it is, you won’t know to look for that spacer — and you’ll be pulling the pump apart again within the hour.
Pusher vs Non-Pusher
Next, look at how the seal maintains face contact. Pusher seals use visible springs — either a single coil spring, multiple small springs arranged in a circle, or a wave spring — to press the faces together. An O-ring or similar secondary seal slides along the shaft as the faces wear.
Non-pusher seals replace the sliding secondary seal with bellows. Rubber bellows look like corrugated accordion tubes. Metal bellows are thin, welded metal convolutions. The bellows element seals — and in metal bellows also provides the spring force — so there’s no sliding O-ring on the shaft. Single coil springs handle roughly 80% of standard water service applications, so that’s the most common configuration you’ll encounter.
Identifying Seal Type Numbers
With spring type and mounting style established, you can narrow the seal to a type designation. Type 1 seals are elastomer bellows designs with a single spring. Type 2 seals are compact non-pusher elastomer bellows designs with a single spring. Type 21 seals are the most common single-spring elastomer bellows design. Each type has different measurement reference points, which is why configuration identification must come before you reach for the calipers.
Take the Eight Critical Measurements
Dimensions are where most identification efforts fail. Your seal supplier needs eight measurements to cross-reference a replacement, and getting any of them wrong by more than 0.030 inches can mean a seal that won’t fit or leaks immediately.
Grab a digital caliper (accuracy to 0.02 mm) and work through these in order:
- Shaft or sleeve O.D. — Measure the shaft or sleeve where the seal rides, not the bore of the seal head. The most common mistake I see is measuring the seal’s inside diameter instead of the actual shaft. Use an outside micrometer for shafts over 50 mm — calipers alone aren’t precise enough.
- Seal head O.D. — The outside diameter of the rotary unit. This must clear the stuffing box bore.
- Seat O.D. — Outside diameter of the stationary seat or mating ring. Measure at the widest point including the gasket if still attached.
- Seat thickness — Axial thickness of the stationary seat. Use a depth gauge for accuracy.
- Stuffing box bore — Internal diameter of the pump’s stuffing box cavity. Measure directly on the pump, not on the old seal.
- Stuffing box depth — Distance from the stuffing box face to the shaft shoulder or seal chamber back wall.
- Spring free length — Uncompressed length of the spring. Stretch it gently to free length — don’t guess from the compressed state.
- Installed working length — Total axial length of the seal assembly when properly set. This is the most critical and most frequently botched measurement.
Type-Specific Adjustments
Here’s where it gets tricky. Measurement procedures change depending on the configuration you identified in the previous section.
For rubber bellows seals (Type 1), add 0.016 inches to the bellows I.D. measurement to account for compression. Skip this adjustment and the replacement won’t fit the shaft correctly.
For operating height on certain bellows types, multiply the total assembled length by 0.730 rather than measuring directly. The compressed bellows changes the effective height, and a straight measurement gives you the wrong number.
These adjustments are why “close enough” doesn’t exist in seal identification. A tolerance of plus or minus 0.030 inches on working length is all you get — outside that window, the seal either binds or has insufficient face loading.
Identify Face and Elastomer Materials
With configuration and dimensions recorded, now turn to materials. Start with the seal faces — the flat, lapped surfaces where the rotary and stationary components meet.
Seal Face Identification
Color is a starting clue, not a conclusion. Black usually means carbon graphite. White or cream means ceramic (aluminum oxide). Gray is where most technicians get tripped up — both silicon carbide (SiC) and tungsten carbide (TC) appear gray.
Pick up both pieces. Tungsten carbide has a density around 15.6 g/cm3 while silicon carbide sits at roughly 3.2 g/cm3 — nearly a five-times difference. A TC face in your hand feels noticeably heavier than SiC of the same size. That weight-in-hand test is more reliable than squinting at shades of gray, especially on worn or stained faces.
One more clue: if the old face shows concentric lapping marks and appears to have been resurfaced, it’s almost certainly a hard face material. Carbon faces typically aren’t re-lapped after service.
Elastomer Identification
Check all the rubber components — shaft O-rings, bellows, seat gaskets. Black elastomers are trickiest because Buna-N, EPDM, and Viton all come in black. If you know the application temperature runs above 325 F, the elastomer is likely Viton or Aflas. EPDM handles steam and water well but swells in hydrocarbons.
The safest approach for elastomer identification is recording the application conditions — fluid type, temperature range, pressure — and letting your supplier match the elastomer. Don’t guess on chemistry.
Decoding the 5-Character Material Code
If your old seal carries a stamped material code (e.g. `BCFZF`, `VCFZF`), it likely follows the 5-position American code; European brands use a different letter key, so confirm the system first. Master this and you can read 80% of any American-code part number without a cross-reference chart.
| Position | Component | Common codes |
|---|---|---|
| 1 | Elastomer / O-ring | B = Buna/NBR (≤225°F, oils/fuels) · V = Viton/FKM (≤400°F, chemicals) · Q = EPR/EPDM (hot water, steam ≤300°F) · T = PTFE/Teflon (universal chemical) · N = Neoprene |
| 2 | Rotary face | Typically the softer material (often carbon) |
| 3 | Metal hardware | Stainless grade — F usually = stainless steel |
| 4 | Stationary face | Typically the harder material (SiC, TC, ceramic) — Z = tungsten carbide |
| 5 | Spring material | Stainless or Hastelloy variant |
A `BCFZF` reads: Buna O-rings / Carbon rotary / stainless steel hardware / tungsten carbide stationary / stainless spring — fine for water and light hydrocarbon service. `VCFZF` swaps Viton for Buna, raising the chemical and temperature envelope. Codes that put a hard material in position 2 indicate hard-on-hard face pairing for abrasive duty. If the code on the seal you pulled does not match the application, you have an installation history to investigate before reordering.
The 3 Identification Mistakes That Lead to Wrong Replacements
After years of helping technicians identify seals over the phone and at the workbench, the same three errors keep coming back.
Mistake 1: Measuring the seal bore instead of the shaft O.D. A worn seal’s inside diameter is always slightly larger than the shaft it rode on. Measure the shaft directly. If the shaft shows wear, measure at an unworn section and note the wear.
Mistake 2: Confusing SiC and TC by color alone. I’ve seen purchase orders come through specifying silicon carbide when the application clearly had tungsten carbide — and the failure happened within days because the material pairing was wrong. Always use the weight test. If the face feels heavy for its size, it’s TC.
Mistake 3: Ignoring the seal arrangement. A single seal and the inboard seal of a dual arrangement can look identical on the bench. But if you order a single seal to replace what was actually a dual inboard, the piping connections and gland configuration won’t match. Before you pull the seal, note whether there’s a second set of faces, a barrier fluid connection, or a tandem/back-to-back arrangement.
What to Record Before Contacting Your Supplier
Seal suppliers don’t need a part number to cross-reference your seal — most identification forms list part numbers as optional. What they do need is everything you’ve just collected: configuration type, all eight dimensions, face materials, elastomer type, and the application data — fluid, temperature, pressure, and shaft speed.
The Three Pieces Most Incomplete Orders Are Missing
Most incomplete orders ship with the same three gaps: seal chamber bore + depth, upset operating conditions, and carbon grade.
Shaft diameter alone does not define seal fit. Three measurements do. The chamber bore determines which stationary seat fits the housing. The chamber depth sets the working length and required spring compression. Get any one wrong and the seal will not seat properly or will not have enough compression to seal. Find both on the pump’s mechanical seal assembly drawing, or measure directly with a depth gauge and inside micrometer once the old seal is removed — clean carbon deposits or scale off the bore first, or you will read a chamber that is too tight. Also record gland bolt circle diameter, bolt count, and bolt size; if the stuffing box face is scored, mention it so your supplier can recommend a different gland plate.
Pressure and temperature need three values, not one. Normal operating (where the seal sits 90% of the time), maximum continuous (highest sustained level during production), and upset/transient (peak during startup, shutdown, or process excursion). Above roughly 14 kg/cm² (200 psi) of seal chamber pressure, balanced seal designs become necessary. Report only normal conditions and your supplier quotes an unbalanced seal that burns out within weeks of the first upset.
Carbon grade specification matters as much as elastomer. If you write “carbon” on the order, you get a generic carbon-graphite face that may fail in hot or oxidizing service. API 682 specifies premium-grade, blister-resistant carbon for a specific reason: lower-grade carbons blister so slowly that the operator attributes rising leakage to normal wear. Pull the API 682 grade if the original seal was running in any thermal or oxidizing duty, and write it on the order explicitly.
Photograph the seal from multiple angles before cleaning it. Decode any visible part number markings — even partial numbers help narrow the search. Record the pump manufacturer, model, and serial number. The more data points your supplier has, the faster and more accurate the cross-reference. If the seal is too damaged to identify through standard methods, a manufacturer with reverse engineering capability can work from your dimensional data and application conditions to produce a match.
