A Gauge That Never Moves, a Channeled Sand Bed, and a Heater That Keeps Tripping

Turnover math: 216,000 gal ÷ 380 gpm = 568 min = 9.5 hours against a 6-hour requirement. To hit 6 hours you need 216,000 ÷ 360 = 600 gpm - a 58% increase. No amount of backwashing produces 58%.

Why the gauge is a problem, not just a reading: a filter gauge is only useful as a differential instrument. You need to see it at clean start, watch it climb as the bed loads, and act at roughly 8–10 psi above clean. A gauge that reads 22 psi forever, and does not drop when you open the air relief, is telling you nothing - and worse, it is telling you that everything is fine. A gauge that does not respond to air relief is plugged at the port or snubber, has a blocked sensing line, or is simply dead. Suspicious corroboration: with a discharge gauge at 20 psi and an influent gauge stuck at 22 psi, the numbers are barely self-consistent.

How to establish loading without trusting it:

1. Fix the instrument. Remove the gauge, clear or replace the snubber and port, install a known-good gauge, open the air relief and confirm it swings to zero, then re-baseline at a clean filter and write the number down.
2. Use the flow trend as the primary signal. Flow down plus pressure up = loading. Flow down with pressure flat or falling = suction restriction, impeller/wear-ring wear, or a discharge-side restriction (including a scaled heater). Your discharge pressure fell from 24 to 20 psi while flow fell 37% - that pattern is not a loaded filter, it is a loss of pump output or an obstruction downstream.
3. Use differential across the filter if you have inlet and outlet gauges, and the sight glass during backwash, as secondary confirmation.
4. Log clean-start pressure and flow after every backwash so the next operator inherits a baseline instead of a mystery.

What the observations mean:

• Milky backwash for two minutes, then clear at 90 seconds: the milkiness is fines and debris being lifted out - expected for the first stretch. The problem is what happens after: if pressure (once you can measure it) climbs back within hours of a backwash, the bed is not retaining fines and flow is short-circuiting.
• Throttled backwash flow so the sight glass does not overflow: this is the root operational error. A sand bed needs roughly 15–20 gpm per square foot to fluidize and expand 15–25%. A throttled backwash never lifts the bed. Debris stays in the upper layer, the media packs and cements, and the bed channels. If the sight glass cannot handle the correct rate, the sight glass plumbing is undersized or obstructed - fix that, not the rate.
• Calcified top 4 inches: calcium carbonate cementation, typically from pH/alkalinity excursions and long intervals between backwashes. Cemented media cannot be cleaned by backwashing at any rate.
• Visible cracks and a 3-inch depression near the inlet: textbook channeling. Water takes the crack and bypasses the bed; the depression is media that has migrated or been lost, and it also signals you should inspect the laterals and nozzles for breakage - media loss into the pool or into the sight glass is the tell.

Decision: replace the media, do not top it up. Eleven-year-old sand with calcification plus channeling plus a disturbed bed is a failed bed. Tapping in a few bags of sand over a cracked, cemented surface produces a filter that looks serviced and filters worse.

Replacement scope: remove all media, inspect and pressure-test the laterals and nozzle assembly (replace cracked laterals), verify the specified media - typically 0.45–0.55 mm effective size - and the correct bed depth per the manufacturer, then re-baseline clean-start pressure and flow, and record the installation date where the next operator can find it.

The relationship: the heater sees the same 380 gpm the filters do - or less, depending on where it sits in the loop. Low flow through the exchanger means a high temperature rise across the heater, a hot outlet, and a high-limit switch doing exactly its job. Two things make it worse and they compound:

• Scale. High pH, high alkalinity, and elevated surface temperatures deposit calcium carbonate on the exchanger. Scale insulates, so the metal runs hotter for the same heat input, the temperature rise climbs further, and the high-limit trips more often. Once it starts, it accelerates.
• Sooting / poor combustion if the burner is also running hot and rich - check the stack and the exchanger for soot on inspection.

Order of repair - restore flow before you touch the heater:

1. Fix the gauge so you can actually see what the filter is doing.
2. Fix the backwash procedure and the bed (previous task) to recover flow.
3. Confirm actual flow through the heater against the manufacturer's minimum - typically 20–40 gpm depending on model - and check that the internal bypass and flow switch are functioning and not stuck.
4. Then service the heater: descale the exchanger, inspect for soot, verify the flow switch and high-limit operation, and confirm the chemical injection point is downstream of the heater. Injecting concentrated chlorine or acid upstream of a gas heater is a corrosion and scale shortcut you will pay for twice.
5. Verify the temperature rise across the heater at design flow before returning to service.

What I refuse to do: I will not bypass, jumper, or repeatedly reset the high-limit switch to keep the pool warm during peak hours. The high-limit is the last protection between a low-flow condition and a failed exchanger, a dry-fire, or worse - and a $6,000–$10,000 exchanger replacement plus downtime is the cheap outcome.

Interim operating decision: with turnover at 9.5 hours against a 6-hour requirement, this pool is out of compliance regardless of the heater. Record the readings, notify the responsible party in writing with the flow deficit and the repair plan, and reduce bather load or close per the code until design flow is restored.