You are here: Home » News » Technical Article » Valve Spare Parts Strategy: Which Kits To Stock, How Many, And Why The Percentage Rule Fails

Valve Spare Parts Strategy: Which Kits To Stock, How Many, And Why The Percentage Rule Fails

Views: 0     Author: J-VALVES     Publish Time: 2026-11-14      Origin: Site

Inquire

facebook sharing button
twitter sharing button
line sharing button
wechat sharing button
linkedin sharing button
pinterest sharing button
whatsapp sharing button
sharethis sharing button
Valve Spare Parts Strategy: Which Kits To Stock, How Many, And Why The Percentage Rule Fails

Almost every valve order ends with the same conversation. The buyer asks what spare parts should be included, the supplier proposes a percentage of the order value, and both sides agree on a number that nobody can defend a year later - when the storeroom holds four complete DN80 ball valves nobody needed and no seat kit for the DN200 trunnion valve that has been passing for six weeks. The problem is not that spares are ignored. It is that they are bought as a percentage of capital spend rather than as a decision about failure modes, lead times and consequences.

This article builds a spare parts plan the way a reliability engineer would: start from what actually breaks, group the population into interchangeable groups, decide which spares tiers the project needs, then use the Poisson distribution to turn a failure rate into a stock quantity with a stated confidence. It ends with the commercial side - shelf life versus stock horizon, repair versus replace, documentation of replacement pressure parts, and the clauses worth putting in the purchase order. Two rules run through everything below: stock what fails, not what costs the most, and the quantity is a calculated number with a confidence level attached, not a percentage of the invoice.

1. Three ways spare parts plans fail

The failures are predictable, and they repeat across projects because they come from how the list is generated rather than from the valves themselves.

The percentage-of-order rule. A buyer is offered "two percent of order value in spares" and signs. The mechanism has no relationship to failure rate, duty cycle, lead time or criticality. Two identical plants with the same valve count but different cycling rates, different media and different supply chains get the same spare package, and the money lands on whichever item happens to be most expensive per unit - usually a complete valve - instead of on the seals and seats that will actually be replaced.

Buying valves instead of parts. A complete spare valve feels like the safest answer, and for a handful of critical long-lead items it is. Applied across a population it is the most expensive possible answer: a complete valve ties up ten to thirty times the capital of the soft goods kit that would have restored it, and it sits on a shelf while the parts that are actually consumed - packing, body seals, seat inserts - are still not in stock.

Ignoring the lead-time clock. Stock levels are usually set against the failure rate alone, with the implicit assumption that a replacement arrives when needed. For a stocked elastomer, that is true within days. For a bespoke seat ring in a duplex or Monel trim, a replacement set may be 12 to 20 weeks out, and the spare quantity has to cover the whole exposure window, not the failure rate on its own.

There is a fourth, quieter failure worth naming because it is invisible in the storeroom: spares that are identified wrongly. In one six-year survey of valve failures in light-water nuclear plants, roughly 11 percent of failures were classed as procedural - the direct result of an error in construction, operation or maintenance - and improperly installed gaskets or parts sat inside that category. A kit that is 90 percent right is not a spare; it is a scheduled rework.

2. What actually breaks: follow the failure data, not the price list

The same survey classified 138 valve failures and found leakage dominant, with operating malfunction second:

Failure category

Share of valve failures

What sits inside the category

What that means for spares

Excessive leakage

44 percent

Of the leakage failures, about 49 percent were stem packing and about 48 percent were leakage past the disc seating surfaces

Soft goods and seat/seal kits carry the demand; complete valves do not

Operating malfunction

37 percent

Loss of ability to perform the intended function, or initiation of an unintended one - actuator, stem, obstruction, control accessories

Actuator seal kits, solenoids, positioners, limit switches

Procedural

11 percent

Wrong gasket or part installed, incorrect setting after maintenance

Identification discipline: tag to bill of materials, revision control

Structural

8 percent

Pressure-retaining parts and load-transmitting parts

Inspection and NDT program, not inventory

Read the first row twice, because it decides where the money goes. Two components - the stem seal and the seat - account for the overwhelming majority of leakage failures, and leakage is itself the largest single failure category. In the isolation-valve subset of that same survey, stem packing leakage alone accounted for about 40 percent of all isolation valve failures. Other published industry breakdowns put seat leakage at roughly 35 to 40 percent and packing at 25 to 30 percent of failures, which is the same conclusion reached by a different route: the parts you consume are the parts that touch the process and move.

Now compare that with a typical first spare parts list, which usually contains one complete valve per size. The economics are upside down. A DN150 Class 300 soft-seated ball valve might be USD 3,000 to 5,000; the seat-and-seal kit that restores it might be USD 900 to 1,500. For the price of one complete spare valve you can hold kits for the three sizes that actually fail, plus actuator seal kits for the two actuator models on site - and cover far more of the failure distribution.

The counter-case matters too: a complete spare valve is the right purchase where the body itself is at risk (aggressive service, severe cycling, weld-end valves in a line that cannot be cut) or where the lead time for a replacement exceeds the tolerable outage. That is a decision about consequence, and Section 8 handles it explicitly.

3. Define the interchangeable group before counting anything

Spare quantity arithmetic only works if the population is grouped correctly, and this is where most lists quietly go wrong. A "spare" is a part that can be installed on a valve without changing its rating, its seal material, its temperature range, its bore, or its approval. Anything that fails one of those tests belongs to a different group and needs its own count.

Four grouping traps appear repeatedly in valve spares:

  • Floating versus trunnion. A DN150 floating ball valve and a DN150 trunnion ball valve do not share seats, bearings or stem parts. Same size, same class, same face-to-face - different internals.

  • Upstream versus downstream seats. In a trunnion valve the two seat assemblies are often not identical, because one may be single-piston-effect and the other double-piston-effect. Ordering "two seats" without specifying position is how a DIB-1 valve gets rebuilt with the wrong seat in the wrong end. The seat geometry and piston-effect logic are covered in the ball valve seat design guide; the spares consequence is simply that the bill of materials must be position-specific.

  • Revision drift. The same model number sold over ten years may have had three stem seal designs and two seat insert materials. A spare ordered from the model number alone can arrive as the current revision and not fit the 2014 valve in the line. The tag number, serial number and the original order reference are the keys that resolve this - which is why the spares list should be built from the as-built bill of materials, not from the catalogue page.

  • Trim and material class. A kit for a standard WCB/13Cr valve is not a kit for the same valve in duplex with hardfaced trim, and it is certainly not a kit for a NACE service valve where the hardness limit applies to the replacement part. Replacement trim has to match the original specification, not just the geometry.

The practical output of this section is a spreadsheet with three columns: interchangeable group identifier, installed quantity in that group, and the bill of materials with revision. Every number later in this article is computed per group, never per plant total.

4. The four spares tiers a project actually needs

Project practice has settled on a tier structure, and it is worth naming because the tiers answer different questions and should be bought at different times.

Tier

When it is bought

Quantity basis

Typical content

How it fails

Commissioning spares

With the valve order

About 2-5 percent of the interchangeable group, minimum one kit per group

Soft seals, packing, gaskets, small operator parts lost or damaged during installation, hydrotest and startup

Left in the site container, never transferred to the maintenance store, re-ordered next year

Operating or two-year spares

With the order or at handover

About 5-10 percent of the group over the two-year window, adjusted for duty

Seat and seal kits, stem packing sets, body gaskets, actuator seal kits, consumable accessories

Counted twice when the commissioning package is rolled into it

Capital (insurance) spares

Project decision, usually after a criticality review

One per selected critical group, not a percentage

Complete valve, or complete trim set for a long-lead bespoke item

Bought for every size, so the critical item is still not covered

Shutdown spares

Per turnaround scope

Scope-driven, per planned overhaul

Overhaul kits, trim sets, actuator overhaul kits, bolting and gaskets for the planned work

Ordered late, so the kit misses the window and the valve is rebuilt with reused parts

Three disciplines make this structure work. First, the percentages are purchasing allowances, not requirements of any standard - no valve standard, API or ISO, prescribes a spare parts percentage for a project, and a supplier who claims otherwise is describing a commercial convention. Second, the commissioning package and the two-year package must be listed separately and the unused commissioning parts credited against the next order, otherwise the plant buys the same seals twice. Third, a minimum of one kit per interchangeable group applies regardless of the percentage, because one percent of a group of six valves rounds to nothing, and the six valves still fail.

Where the installed population is large enough to be counted in dozens per group, convert the percentage to a discrete quantity rather than leaving it as a number to be interpreted at the packing list:

Valves in one interchangeable group

Commissioning kits

Two-year kits

1-10

1

1

11-25

1-2

1-3

26-50

1-3

2-5

51-100

2-5

3-10

101-200

3-10

6-20

Use the lower end of each range for clean, low-cycle service with fast local supply, and the upper end for dirty service, frequent cycling, special seal materials, remote sites or long delivery. Those two variables - duty and access - move the number by a factor of two or three, which is far more than the difference between a 2 percent and a 5 percent rule.

5. What is inside a kit, valve by valve

The word "kit" is used loosely in the trade, and the ambiguity costs money: a seat-and-seal kit, a soft goods kit and a full overhaul kit are different products with different contents, and buying all three for the same valve buys the same seals twice. Define contents on the order.

Valve type

Soft goods kit

Overhaul or revision kit adds

Notes

Ball valve, floating

Seat rings, body seal or O-ring, stem seal, thrust washer, anti-static device if separate

Ball, stem, seat retainer or end cap hardware

Seat material must match the original pressure-temperature envelope; a PTFE kit in a PEEK application will pass the test and fail in service

Ball valve, trunnion

Two seat assemblies (position-specific), body gaskets, stem seal set, trunnion bearings and thrust bearing

Ball, stem, seat springs, cavity relief components

Upstream and downstream seats may differ; bearings are a wear item that is often forgotten and drives torque

Gate valve

Stem packing set, bonnet gasket, gland and eye bolts if corroded

Wedge or gate, seat rings for the renewable-seat designs, stem, backseat bushing, stem nut

Wedge configuration (solid, flexible, split) must match; see the gate valve wedge design guide

Globe valve

Disc and seat ring, stem packing, bonnet gasket

Stem, plug, cage or throttle trim, backseat

Disc and seat are a matched pair; replacing one against a worn one re-creates the leak

Butterfly valve

Seat liner or seat ring, shaft seals, bushings, end-of-shaft cover gaskets

Disc, shaft, taper pins, body

Liner material and the seat-to-disc interference set the torque; a wrong liner changes the actuator requirement

Check valve

Disc or plate, hinge pin, seat ring, cover gasket

Full internals, spring

Spring rate and disc weight set the closing behavior; substituting changes water hammer performance

Pneumatic actuator

Piston or diaphragm seals, O-rings, guide rings, lubrication

Springs, shaft, bearings, end caps

Spring packs are matched to the fail-safe torque; see the pneumatic actuator failure mode guide

Two contents details deserve explicit lines in the purchase order because they are routinely assumed rather than stated. First, whether the stem seal kit is inside the seat-and-seal kit or separate - if the plant cannot tell, it will discover the gap during a shutdown. Second, whether fasteners, gland bolts, gaskets and lubricant are included: a kit that arrives without the body gasket is a kit that stops the job.

6. How many: the Poisson arithmetic

With the group defined and the kit contents agreed, the quantity is a probability question. Failures of a maintained population arrive independently at a roughly constant average rate during useful life, so the number of failures in a window follows a Poisson distribution. Given the expected number of failures in the exposure window, the coverage of holding n spares is the cumulative probability:

P(X <= n) = sum over k = 0 to n of [ e^(-L) x L^k / k! ], where L = lambda x T is the expected number of failures in the window, lambda is the failure rate per valve per year, and T is the exposure window in years.

The exposure window is the part that is usually got wrong. It is not one year and it is not the maintenance interval: it is the period during which the plant is exposed if the shelf is empty, which is the replenishment lead time plus the time to discover and raise the order. Stocking to cover a year of failures when the part arrives in three weeks over-buys by an order of magnitude; stocking to cover a three-week lead time when the part is a 16-week bespoke casting under-buys catastrophically.

Expected failures in the window (L)

Stock for 90 percent coverage

Stock for 95 percent

Stock for 99 percent

0.03

0

1

1

0.2

1

1

2

0.5

1

2

3

1.0

2

3

4

2.1

4

5

6

3.2

5

6

8

6.0

9

10

12

The table carries two lessons that a percentage rule cannot express. The first is that coverage is expensive at low failure counts: at L = 1.0, going from 90 percent to 99 percent coverage costs two more kits for a nine percentage point gain, while at L = 6.0 the same jump costs three kits for eight points. The second is that the quantity scales with the square root of the failure count, not linearly with it - doubling the population does not double the spares, which is why a plant with 300 valves of one type does not need twice the kits of a plant with 150.

Worked example. A terminal has 300 soft-seated ball valves of one interchangeable group, spread over three sizes that share seats and stem seals. Maintenance records show 12 seat or packing interventions in the last 12 months. The replenishment lead time for a kit is 14 weeks, and add two weeks to raise the order: 16 weeks, or 0.31 years.

The failure rate per valve-year is 12 / 300 = 0.04. The expected failures in the exposure window are L = 300 x 0.04 x 0.31 = 3.7. Reading the table at L = 3.2 and adding the difference for 3.7, six kits gives roughly 93 percent coverage and eight gives about 99 percent. Note what the arithmetic did not use: the valve price, the order value, or the number of valves on its own. It used the observed failure rate and the supply chain, which are the only two things that determine whether the shelf is empty when a valve starts passing.

A word on where lambda comes from. Plant maintenance records are the best source, because they include installation damage, wrong parts and operator error - the real distribution. Vendor MTBF figures are the worst source in isolation, because they are generated under idealized conditions; where nothing else exists, use them with a derating factor in the 0.3 to 0.7 range and revisit the number after two years of site data. A spares list built on a vendor MTBF taken at face value will be systematically short.

7. Where to stop: the marginal cost of the last kit

The Poisson table answers "how many for 95 percent", but not "should I pay for 95 percent". That is an economic question, and it has a clean answer: keep adding kits while the annual cost of holding one more is less than the expected loss it avoids. The holding cost of a spare is not its price - it is the capital tied up plus storage, insurance, handling and obsolescence, commonly taken as 20 to 30 percent of part value per year. Use 22 percent as a working figure.

Continuing the terminal example at L = 3.7, with a kit price of USD 1,450 and a consequence of USD 24,000 for an unplanned 6-hour outage (6 hours at USD 4,000 per hour of lost margin):

Kits held

Coverage

Stockout risk

Holding cost per year

Margin gained over previous row

Expected loss avoided

Verdict

4

about 77 percent

23 percent

USD 1,276

-

-

Under-covered

5

about 89 percent

11 percent

USD 1,595

11.7 points

USD 2,808

Clearly worth it

6

about 95 percent

4.7 percent

USD 1,914

6.3 points

USD 1,510

Worth it

7

about 98 percent

1.8 percent

USD 2,233

2.9 points

USD 699

Worth it

8

about 99.4 percent

0.6 percent

USD 2,552

1.2 points

USD 283

Not worth it

The curve has an answer: seven kits, and the eighth is a donation. This is the argument that wins budget reviews, because it converts an inventory policy into a number with a stated consequence behind it. It also explains a pattern that puzzles buyers - why a supplier can be right about the need for spares and wrong about the quantity at the same time. The need follows from the failure distribution; the quantity follows from the failure rate, the lead time and the cost of the outage, none of which the supplier knows.

Two caveats keep this from becoming a false precision. The consequence figure dominates the answer: if the outage costs USD 120,000 rather than USD 24,000, the eighth kit is worth it and the tenth may be too, and the decision moves. And the model assumes independence - it does not know that a bad batch of packing will fail ten valves in the same month, or that a commissioning error will take out every kit in the store. Where common-cause failure is plausible, add a buffer outside the arithmetic rather than pretending the Poisson distribution covers it.

8. Insurance spares: when probability is the wrong argument

For a single critical valve, the probability arithmetic produces a number that looks like an argument against stocking. Take a DN600 Class 900 trunnion ball valve on a pipeline that cannot be bypassed, with a 30-week lead time for a replacement (0.58 years) and a soft goods failure rate of 0.05 per year. Then L = 0.05 x 0.58 = 0.029, and the probability of no failure at all in that window is already 97.1 percent. One kit takes it to 99.96 percent.

So the kit buys 2.9 percentage points of risk reduction, and if the outage costs USD 180,000, the expected value of stocking it is about USD 5,200 - less than the USD 9,000 the kit costs. On expected value the kit loses. It should still be bought, and the reasoning has to be stated honestly: the decision is driven by the tail, not the mean. Without the kit, a seal failure in month two means either a bypass that does not exist, a 30-week wait, or a temporary repair on a Class 900 joint. The expected-value arithmetic assumes the plant can absorb the bad outcome at its average cost, and for a single-point-of-failure item it cannot.

The same logic explains the second insurance decision, which is whether to hold hard parts. The failure rate for a ball, wedge or stem is roughly an order of magnitude below the soft goods rate - for the example valve, L = 0.005 x 0.58 = 0.003, so a stock of zero already covers 99.7 percent of windows. Holding a spare ball is therefore not risk management; it is either a response to a specific known mechanism (erosion in slurry service, galling in a stainless pair, cavitation damage downstream of a letdown) or a response to a lead time so long that the part cannot be re-made inside any acceptable outage. Say which one it is, in writing, or the part will sit on the shelf for fifteen years and then be scrapped at the next inventory purge.

In practice the critical spares list for a plant or a platform is short - typically a handful of items - and it is built from two questions asked of every tag: what happens if this valve cannot be isolated, and how long does the replacement take to arrive. A positive answer to the first combined with a long answer to the second is the whole case.

9. Shelf life versus stock horizon

Elastomer spares age on the shelf, and the clock starts at cure, not at delivery. ISO 2230 assigns shelf lives by polymer family, and the practical consequence is that a "ten-year strategic stock" of seals is a fiction for most materials:

Elastomer family

Typical shelf life per ISO 2230

Where it goes wrong in a valve store

NR, SBR

about 5 years, extendable to 7 after inspection

Rare in process valves, common in general gaskets

NBR, CR, ACM

about 7 years, extendable to 10 after inspection

Nitrile O-rings and seat inserts in unairconditioned stores in hot climates

EPDM, FKM, FFKM, VMQ

about 10 years, extendable to 15 after inspection

FKM is the common seat and body seal material in hydrocarbon service, so the 10-year figure is the one that matters most

PTFE and filled PTFE

No comparable shelf-life limit; degradation is by creep and contamination rather than aging

Deformation under its own weight if stored badly; seals flattened in a box

Graphite packing and gaskets

Effectively unlimited if kept dry

Contamination and moisture pickup, not age

Three rules follow. Stock to the failure rate and lead time, not to a target year - the arithmetic in Section 6 already produces the right horizon, and it is usually months, not decades. Apply first-in-first-out by cure date, and record the cure date on the kit label, because a kit with no date is a kit of unknown age. And when a seal passes its shelf life, inspect before scrapping: the criteria are cracks, hardening, tackiness, and compression set - a part that fails them is scrap, a part that passes is usable with a documented extension.

A white bloom on a nitrile part is not a rejection criterion. It is a bloom of compounding ingredients migrating to the surface, it is normal, and it wipes off; rejecting on appearance alone throws away serviceable stock and trains the store to distrust its own inspection. Storage conditions for the rest of the spares package - including the question of whether a spare valve should be stored open or closed - are handled in the valve storage and preservation guide.

10. Lead time is the variable that decides the plan

Failure rate tells you how often; lead time tells you how long you are exposed. In most spares decisions the lead time is the bigger lever, and it is the one that procurement can actually change by paying for stock rather than for production slot.

Spare part type

Typical replenishment lead time

What drives it

Stocking consequence

Standard soft goods kit, stocked by the maker

1-4 weeks

Warehouse location and shipping

Low stock is defensible; rely on supplier stock

Standard soft goods kit, made to order

6-12 weeks

Seal material availability, machining queue

Stock to cover the lead time

Trim set with special alloy or hardfacing

12-20 weeks

Forging or casting, hardfacing, machining

Stock one per critical group; long-lead item

Body, bonnet or stem in a special material

16-30 weeks

Pattern or die, heat treatment, NDT, certification

Insurance tier only

Actuator or control accessory

8-20 weeks

Electronics, approvals, configuration

Stock modules, not complete units

Repair by the original manufacturer

6-14 weeks including transit

Workshop slot, coating, test

Rotation set is cheaper than a spare valve

The numbers are project-dependent, but the shape is stable, and it produces one strong recommendation: ask for the lead time of every line on the spare parts list, in writing, at the time of the valve order. A spares list without lead times cannot be turned into a stock quantity; it is a shopping list. Where a lead time is long and the item is critical, the alternatives to holding stock are a consignment arrangement, a guaranteed exchange pool, or a contractual commitment from the maker to hold semi-finished stock for the project - all three are cheaper than a complete spare valve and all three need to be negotiated before the order is placed, not after the failure.

11. Repair or replace: when a kit is a false economy

A kit restores the sealing function; it does not restore the pressure boundary. Before a repair is approved, four things decide whether the body is worth the kit.

Wall thickness. Ultrasonic measurement at the seat pocket, the body-bonnet joint and the lowest point of the cavity, compared against the minimum wall for the size and class in the applicable standard. A body that has lost metal to erosion or corrosion below the minimum is scrap regardless of how well it seals.

Seat pocket and sealing surfaces. Corrosion or wire-drawing in the seat pocket means a new seat ring will not seal against the body, and the defect is not visible on the bench until the valve is back in line. This is the classic "repaired twice, still passing" signature.

Hardfacing condition. On metal-seated valves, the seat faces are the sealing element. If the hardfacing is cracked, spalled or worn through into the base metal, a new seat ring is the only repair, and re-machining beyond the hardfacing depth removes the very layer that was doing the work.

Threads, guides and bores. Stem threads, the backseat shoulder, the bonnet bore and the guide surfaces take the operating loads. A kit that renews the seals on a stem with worn threads produces a valve that holds pressure and cannot be operated.

Where the body passes, the repair still has to be proven. A repaired valve is a valve whose pressure boundary has been opened, and it goes back through the same shell and seat testing as a new one - acceptance per API 598 pressure testing, with the leakage class stated on the repair order rather than inherited by hope. This is also where the spares decision comes full circle: the seat leakage class the plant actually needs determines whether a soft-seated kit is acceptable after repair, or whether the valve has to come out of that service. The mapping is set out in the valve seat leakage class guide.

12. Documenting replacement pressure parts

A replacement body, bonnet, stem or seat carrier is a pressure-retaining part, and it needs the same material documentation as the original. In practice this is where spares from a second source or from a generic kit fall down: the geometry is right, the certificate is missing.

Three requirements cover most of it. The replacement part should carry a material certificate to EN 10204 type 3.1 at minimum, with the heat number traceable to the mill certificate, and the certificate should state the product form, the heat treatment condition and the mechanical results - not just the grade name. The alloy should be verified on receipt by positive material identification, remembering that handheld XRF cannot read carbon, sulfur or phosphorus; a carbon steel or a low-carbon stainless grade needs optical emission for that. And the part should be marked so that the marking survives installation: a heat number etched on a seat ring is worth more than a certificate in a folder, because it is the only link that survives a decade in a storeroom.

For sour service, hydrogen service or any application where the material specification carries a hardness limit, the replacement part inherits that limit. A kit bought on part number alone does not carry it, and the failure mode - a hard seat ring in a NACE application - is not detectable at commissioning. Specify the service, and ask for the hardness results with the kit.

13. Contract structures that produce spares instead of promises

Spares availability is a commercial construct as much as a technical one, and the clauses that matter are short.

Recommended spare parts list as a deliverable. Require the maker to issue a recommended spare parts list with the order, separated into commissioning, two-year operating and capital items, with part numbers, kit contents, lead times, unit mass, storage requirements and shelf life for each line. This document is the input to every calculation in this article; without it the plant is guessing.

Twenty-year availability, with an escape route. A commitment to supply spares for the life of the equipment is standard language and worth having, but it needs a mechanism: either a last-time-buy notification with a stated notice period, or a documented drawing package that allows a third party to manufacture. Without one of those, the commitment expires quietly when the product line does.

Form-fit-function replacement. Where a maker revises a part, the replacement must be interchangeable without changing the rating, the seal material, the temperature range or the approval - or the change must be notified in advance with the affected serial number range. Revision drift is the single most common cause of a correct part number arriving as an incorrect part.

Consignment and exchange pools. For a large installed base, consignment stock held at site but invoiced on consumption converts inventory into a service, and an exchange pool - where the maker ships a rebuilt unit against the returned core - covers the long-lead items without the plant owning them. Both need a defined replenishment time and a defined condition standard for the returned core, otherwise the pool degrades into a scrap heap.

Price escalation and currency. A ten-year availability commitment with no price mechanism is not enforceable in practice. Agree the basis - index, list price with capped escalation, or fixed for a defined period - at order stage.

14. Failure signatures of the spares program itself

Spares programs fail in recognizable ways, and the signature usually appears in the maintenance record before it appears in the budget.

What you observe

What it usually means

How to verify

Correction

Same valve repaired twice in a year with new seals

Seat pocket or body sealing surface damaged; kit cannot fix it

Measure the pocket, inspect the seat contact band

Replace the body or the valve; stop buying kits for it

Kits on the shelf, valve still waiting

Wrong revision or wrong position (upstream versus downstream seat)

Compare kit part number to the as-built bill of materials by serial number

Position-specific kits, revision-controlled list

Seal failures clustering after a shutdown

Installation damage, wrong lubricant, or aged elastomer issued from the store

Check cure dates and the issue record for that work order

FIFO by cure date, installation training, lubricant compatibility list

Actuator spares unused for years, then three failures in a month

Common-cause: same batch of seals or same air supply problem

Audit instrument air dew point and particulate

Fix the air, not the inventory

Emergency purchases at premium freight every quarter

Stock level set from order value rather than failure rate and lead time

Run the Poisson calculation on the last two years of interventions

Reset quantities per interchangeable group

Complete spare valves in store, no kits

Percentage-of-order purchasing

Compare inventory value by category against consumption

Convert one or two complete valves into kits across the sizes that fail

Spares written off at the inventory purge

Insurance items bought without a stated scenario

Ask which failure each capital spare answers

Keep only the ones with a written scenario and a lead time that justifies them

15. A procurement checklist for the spare parts package

  1. Installed population listed by interchangeable group, with as-built bill of materials and revision reference - not by model number alone.

  2. Recommended spare parts list issued as a deliverable, split into commissioning, two-year operating, capital and shutdown tiers.

  3. Kit contents defined line by line: what is in the soft goods kit, what the overhaul kit adds, whether stem seals and body gaskets are included, and whether fasteners and lubricant are supplied.

  4. Seat position specified for every trunnion valve: upstream and downstream kits listed separately.

  5. Lead time stated per line, and the replenishment time the plant should plan against stated separately from the production lead time.

  6. Failure rate basis agreed: site history where it exists, derated vendor MTBF where it does not, with a review after two years.

  7. Quantities calculated per group from failure rate and exposure window, with the target coverage stated in the document.

  8. Shelf life and cure date marked on every elastomer kit, with storage conditions and a FIFO rule.

  9. Material certification specified for all pressure-retaining replacement parts, to EN 10204 3.1 minimum, with heat numbers marked on the part and PMI on receipt.

  10. Availability commitment with a mechanism: last-time-buy notice period, or a drawing package that permits third-party manufacture.

16. What to do with this on Monday

The plan does not need software. It needs three inputs and one afternoon. Pull two years of maintenance records and count interventions by valve type and size - that gives the failure rate. Ask the supplier for lead times on the ten lines that matter - that gives the exposure window. Put a cost on an unplanned outage of the affected unit - that gives the consequence. The Poisson table converts the first two into a quantity, and the marginal table tells you where to stop.

Then buy accordingly, which in most plants means fewer complete valves and more kits, more actuator seal kits than anyone expected, and a much shorter list of capital spares than the last project carried. If the resulting list looks thin compared with the old one, that is usually the sign it is right: the previous list was priced by the invoice, and this one is priced by the failure.

For the valves themselves, the ball valve product range, the gate valve range and the butterfly valve range can each be quoted with a spare parts package built on the tiers described here, with the kits identified by tag number rather than by model number alone. Related technical articles on seat design, leakage classes, pressure testing, actuator failure modes and long-term storage are collected in the technical article library.

How many spare valves should a plant keep?

Very few, and the logic is not symmetric with the kit logic. A complete spare valve costs roughly ten to thirty times the soft goods kit that restores the same valve, and the failure data says the seals and seats are what fail, not the body. The defensible approach is to stock kits across every interchangeable group that has real consumption, and to hold a complete valve only where the body itself is at risk or where the replacement lead time exceeds the tolerable outage. That usually means a handful of complete valves per plant, selected by asking of each tag what happens if it cannot be isolated and how long a replacement takes to arrive.

What is the difference between a soft goods kit and a full repair kit?

A soft goods kit contains the consumable seals - seat rings or seat inserts, body seals or gaskets, stem packing or stem seals, and usually thrust washers and bearings. A full repair or overhaul kit adds the hard parts: ball, wedge or disc, stem, seat carriers, springs and fasteners. The distinction matters commercially because the names are used loosely and kits overlap - one maker's overhaul kit may already include the seats, body seals and packing that another maker sells as a separate soft goods kit. Specify the contents line by line, and do not count a soft goods kit, a seat-and-seal kit and an overhaul kit as three separate spares until the supplier has confirmed what each contains.

Do valve spare parts have an expiry date?

Elastomer parts do, and the clock starts at cure rather than at delivery. ISO 2230 sets shelf lives by polymer family: about 5 years extendable to 7 for natural rubber and SBR, about 7 years extendable to 10 for NBR, CR and ACM, and about 10 years extendable to 15 for EPDM, FKM, FFKM and silicone. PTFE and filled PTFE have no comparable age limit - they degrade by creep and contamination, not aging - and graphite packing and gaskets are effectively unlimited if kept dry. Two practical rules follow: mark the cure date on every kit and issue first-in-first-out, and inspect rather than scrap when a part passes its stated life. Cracks, hardening, tackiness and high compression set are rejection criteria; a white bloom on a nitrile part is normal and is not.

Does a repaired valve need to be pressure tested again?

Yes. A repair opens the pressure boundary, so the valve goes back through the same shell test and seat test as a new one, with acceptance per API 598 - shell test at 1.5 times the 38 deg C rating, seat test at 1.1 times, and the leakage acceptance stated on the repair order rather than assumed. The seat test deserves particular attention on repaired valves, because the leakage class the plant actually needs decides whether a soft-seated kit is acceptable after the repair, or whether the valve has to come out of that service. Re-testing is also the only way to catch the repair that seals on the bench and passes on the line - a damaged seat pocket that a new ring cannot seal against.

Is it cheaper to hold spare valves or spare kits?

Kits, by a wide margin, for any valve that can be repaired in place or in a workshop. The arithmetic is straightforward: a soft goods or seat-and-seal kit typically costs 20 to 40 percent of a complete valve, and the holding cost of inventory is commonly taken as 20 to 30 percent of part value per year. Four kits across the sizes that actually fail therefore cost less to own than one complete valve and cover far more of the failure distribution. The exception is the single critical item with a long lead time, where the decision is driven by the tail rather than the average: expected-value arithmetic often says do not stock, while the consequence of a thirty-week wait says stock exactly one.

Manufacturer and supplier of industrial valves, including Floating Ball Valves, Trunnion Ball Valves, Flanged Gate Valves, Welded Gate Valves, High Pressure Gate Valves, Globe Valves, Swing Check Valves, Double Disc Wafer Check Valves, Y Strainers , etc. For more information, please send us your email .

QUICK LINKS

PRODUCTS

CONTACT US

Tel:+86 15158426559
Fax:0086 577 56692524
WhatsApp:+86 13600648865
Address:Wenzhou China
Ball Valves,Gate Valves,Globe Valves,Check Valves,Butterfly Valve,Strainer.
Copyright © 2020 J-VALVES. All Rights Reserved | Support by Leadong | Sitemap |  Privacy Policy