What PoolBoss Says
Repair it when the failure is a seal, gasket, gauge, or valve part and the tank itself is sound and under about ten years old. Replace it when the tank is cracked, the same internals have failed twice, or the repair costs more than about half a new filter.
You are standing at the pad with water running down the side of a filter tank and a customer waiting on a number. The part is cheap. The tank is old. That is the whole decision, and you have about two minutes to make it before the conversation turns into a guess.
Most of what is written about filter repair is a parts tutorial - eight steps to change a spider gasket. None of it tells you whether changing that gasket on a 14-year-old tank is worth the callback. What follows is the rule, the five failures that cover almost every filter call, how pressure history tells you which one you are looking at, and what belongs on the quote.
At a glance
Key takeaways
- Repair the part and replace the vessel: seals, gaskets, gauges, and valves are consumables, and the tank is not.
- Repair a tank under about ten years old on its own merits, and treat a second failure in two seasons as a replacement conversation.
- When the repair passes roughly half the installed cost of a new filter, replace it - a new filter also resets the clean-pressure baseline.
- Never patch a cracked tank: it holds 10-25 psi every hour the pump runs, and a patch becomes the weakest point in a pressure vessel.
- Read pressure against the pool's own clean baseline, not a generic number, and service around 8-10 psi above it.
- Pressure that will not come down after a proper cleaning means media or elements; pressure that suddenly drops means something is bypassing.
- Put the repair and the replacement on one quote with a photo of the failure, and it gets answered in one reply instead of two.
Should I repair or replace a pool filter?
Repair the part, replace the vessel. Almost every filter decision collapses into that sentence: seals, gaskets, gauges, valves, and internal elements are consumable parts with a price and a labour time attached, and the tank they sit in is a pressure vessel that either holds or does not. Three thresholds settle which side of the line you are on.
The first is age. A well-maintained cartridge or DE filter commonly runs 10-15 years and a sand tank often longer, so a failure on a tank under about ten years old is usually worth repairing on its own merits. Past that mark every repair is being spent against a shorter and shorter remaining life, and the customer is the one who notices.
The second is repetition. A first failure is a part. A second failure on the same tank inside a couple of seasons is the tank telling you that everything else on it is the same age as the thing that just went. Two repairs on one filter in two years is a replacement conversation whether or not the third failure has happened yet.
The third is cost, and it is the one that ends most arguments. When a repair passes roughly half the installed cost of a new filter, replacement wins on arithmetic alone - and it wins by more than the numbers show, because a new filter arrives with a fresh clean-pressure baseline and no callback risk attached. A $150 reseal against an $1,100 replacement is an easy repair. A $500 multiport rebuild against that same $1,100 is not.
Five failures cover almost every filter service call
The variety on the pad is smaller than it looks. A spider gasket, a tank O-ring or clamp band, a pressure gauge, the internal elements - laterals in a sand tank or grids in a DE tank - and the tank itself account for very nearly every filter call you will run in a season. Four of the five are repairs. One of them is not.
The money spread is wide, which is why the decision is worth two minutes rather than a reflex. A spider gasket is roughly a 30-minute job with a $15-$50 part. Simple seal, O-ring, and gauge work typically lands around $100-$150 all in. Multiport valve rebuilds, lateral replacement, and DE grid assemblies commonly run about $200-$500. A cracked tank or a full internal rebuild reaches $700 or more, against a full filter replacement of roughly $250-$2,000 depending on type and size.
Note where the internal elements sit in that list. Broken laterals and torn DE grids are a repair the first time and a judgement call the second, and they announce themselves as sand blowing back into the pool rather than as anything visible at the pad. Diagnose that symptom properly before you quote against it, because the same debris in the pool can come from a broken standpipe or a bypassing spider gasket, and those are three different prices.
| Failure | Typical cost | Time on site | Verdict |
|---|---|---|---|
| Spider gasket | $15-$50 part | About 30 minutes | Repair |
| Tank O-ring or clamp band | About $100-$150 | Under an hour | Repair |
| Pressure gauge | About $100-$150 with the fitting | Minutes | Repair, every time |
| Laterals or DE grids | About $200-$500 | 1-3 hours | Repair once, weigh it twice |
| Cracked tank | $700+ to rebuild, $250-$2,000 to replace | Half a day | Replace |
A cracked tank is a replacement, not a repair
A filter tank is a pressure vessel, and that is the entire argument. It spends every hour the pump runs holding water at roughly 10-25 psi in normal service, and it is built and rated to survive well above that - which is exactly why a crack in it is not the same class of problem as a leaking seal. A seal that fails weeps. A tank that fails at pressure comes apart, and there is a lid and a clamp band in the immediate area.
Epoxy, putty, and fibreglass patches are sold for this and they do hold, briefly. What they cannot do is restore the wall strength of the original moulding, so the patch becomes the new weakest point in a vessel that cycles up to pressure and back down every single day the system runs. A repair that is fine at rest and marginal at pressure is not a repair on a component like this. It is a delay with your name on it.
The version to give the customer is short and it does not require them to trust you on faith: this is the one part of the system that is holding pressure all day, a patch on it is a temporary fix rather than a repair, and if it lets go it does so with force. Most customers who push back on a $1,100 filter accept it within about thirty seconds of hearing that sentence, because they were arguing with the price rather than with the physics. Say it once, plainly, then price it.
Pressure history tells you which failure you are looking at
Two numbers separate almost every filter diagnosis: the pool's own clean baseline and what it is reading today. A filter reads its clean pressure right after a proper cleaning, and it is due for service around 8-10 psi above that. Those figures are per pool, not universal - a 14 psi baseline on one system and a 22 psi baseline on another are both fine, and comparing either against a generic number tells you nothing.
Pressure that climbs and comes back down when you service it is a filter working correctly. Pressure that will not come down after cleaning it properly first is the diagnostic fork: on a DE or cartridge tank it usually means the media is spent or the elements are loaded past recovery, and on sand it often means the bed has channelled or hardened. That is a media or element problem, and it is where the repair-or-replace question actually starts.
Pressure that drops suddenly is the more urgent signal and it points the other way. A filter that was running at 20 psi and now reads 12 psi with the same pump has usually not improved - something is letting water past rather than through, which means a bypassing valve, a failed internal, or a leak. Sudden low pressure and debris returning to the pool in the same week is close to diagnostic on its own.
None of that reads as a pattern from a single visit, which is the practical problem with diagnosing on the spot. Three weeks of a pool sitting 8 psi over its own baseline is a clear story; the same reading seen once is just a number. When the filter pressure you logged at the last visit sits next to today's reading in the pool's own record, along with what the tech did about it, the trend does the diagnosis for you and the quote stops being a judgement call made from memory.
What goes on the quote so it gets approved once
Price both options, on one quote, at the same time. The repair-or-replace decision is not actually yours to make on a customer's equipment - yours is to make it decidable, and a quote with a single number on it forces the customer to either accept it or start a second conversation. A quote with the repair and the replacement priced side by side, with the age of the tank stated as a plain fact, gets answered in one reply.
An operator running 105 pools across Ocala and The Villages gets this call on a Tuesday. The Ocala pool's sand filter has run 22 psi against its own 14 psi clean baseline for three weeks and it has now started weeping at the clamp band. The band and O-ring are a $40 part and a 30-minute job, so the repair is genuinely easy - except the tank is 14 years old and this is its second leak in two seasons. He sends one quote with both lines on it: $150 to reseal it now, or $1,100 to replace the filter and reset the baseline. The customer picks the replacement, and picks it that afternoon, because the alternative was written down next to it instead of being hinted at over the phone.
The other half of a quote that lands once is evidence. A photo of the weep, the tank's date, and the three pressure readings that led to the recommendation turn a price into a finding, which is most of what belongs on a repair quote in this trade. A customer who can see the failure argues about the number far less than one who is being told about it.
The last thing is timing, and it is a routing problem rather than a sales one. A clamp band that has started weeping in July is a scheduled repair; the same band discovered in August because nobody looked is an emergency call on a Saturday. Making the filter a fixed stop on the pad check that catches it early is what converts these into quotes you send on your own schedule instead of callbacks that pick the schedule for you.
FAQ
Frequently asked questions
How long should a pool filter last before it needs replacing?
Plan on 10-15 years for a cartridge or DE filter and often longer for a sand tank, assuming it has been cleaned on schedule and not run at pressure with a blocked element. The number that matters more than the calendar is how the tank has been treated: a filter that spent five summers running 12 psi over its clean baseline has been stressed in a way a well-serviced tank of the same age has not. Sand itself is a separate clock - the media is usually changed every 5-7 years, and that is a media job, not a filter replacement. Use age as a threshold rather than a countdown. Under about ten years, repair the failure in front of you. Over it, price the repair against a replacement every time, because the next failure is rarely far behind and the customer will remember that you charged for both.
Can I patch a cracked filter tank as a temporary fix?
You can, and it will hold for a while, but it should not go out as a repair. A filter tank holds roughly 10-25 psi every hour the pump runs, and an epoxy or fibreglass patch does not restore the original wall strength - it becomes the weakest point in a vessel that cycles to pressure and back down daily. If it lets go, it does so with force, near a lid and a clamp band. If a customer genuinely cannot replace the filter this week, the honest version is to say the patch is a stopgap to get through a specific short period, write that on the invoice in those words, and quote the replacement at the same time. What you should not do is take money for a patch, call the job closed, and leave the liability sitting on someone else's pad with your company name attached to the last service on it.
What does it mean if filter pressure stays high right after cleaning?
It means the problem is the media or the elements, not the dirt you just removed. A correctly cleaned filter should return close to that pool's own clean baseline within a few minutes of restarting - if it comes back up to 8 psi over baseline immediately, the cleaning did not restore flow. On a cartridge, that usually means the pleats are permanently loaded with oils, minerals, or old algae and the cartridge is at the end of its life. On DE, it points to grids that need stripping and recharging or fabric that has failed. On sand, it commonly means the bed has channelled or hardened into clumps, which a backwash will not fix. Check the obvious first - a closed valve, a clogged impeller, or air in the system all read as pressure oddities - and then treat it as a media replacement rather than another cleaning.
Should I quote a filter replacement in the same size or go bigger?
Go bigger where the pad and the budget allow, because an oversized filter is one of the few upgrades that pays the customer back every week rather than once. A larger filter area means lower pressure per square foot, longer runs between cleanings, and less strain on the pump - and it costs you less service time on that pool for the rest of the time you own the route. Matching the old size is only the right call when the pad space genuinely will not take more, when the plumbing would need reworking to fit it, or when the customer is selling the house. Size against the pool volume and the pump's flow rate, not against the tank that happens to be sitting there, because the original was often specified to a builder's budget rather than to the pool.
Do I need to carry spider gaskets on the truck?
Carry them if you run more than a handful of multiport valves on your route, because it is a $15-$50 part that converts a return trip into a 30-minute fix on the spot. The economics are not close: one avoided second visit covers the cost of a small stock of the common sizes several times over, and the customer experience of a leak fixed the same morning is worth more than either. The practical approach most operators land on is to stock the two or three gasket sizes that match the valves actually on their route rather than a universal kit, along with tank O-rings and a couple of pressure gauges, which fail constantly and cost almost nothing. Note what you use as you use it, so the truck stock reflects the route rather than what happened to be on sale.
How do I explain a $900 filter replacement to a customer who expected a repair?
Lead with the thing that is not repairable, not with the price. The sentence that works is that the tank itself has failed rather than a part on it, that it holds pressure every hour the pump runs, and that a patch on a pressure vessel is a delay rather than a fix. Then show the repair number next to it so the customer can see you are not simply upselling - a $150 reseal on a 14-year-old tank is a real option and it deserves to be priced honestly, including the part where it may not last the season. Add a photo of the crack or the weep and the last few pressure readings that led here. Most of the resistance to a replacement number is not about the money, it is about not being able to see what you are seeing, and a photograph solves that faster than any explanation.
