KEYSTONEIndustrial Services
Interactive trainer

DP & hydrostatic level

A hydrostatic level transmitter never sees the level — it senses the head pressure the liquid puts on a tap, and head equals level times density. Tell it the wrong specific gravity and every reading is scaled off; lose a wet-leg fill, plug a tap, or skip the zero and it lies in its own way. Six tanks, four reads — indicated level vs a dip, head, legs/taps, SG and zero — including the two pairs that look identical until you check the right thing.

By Francisco Padilla Interactive Technician level

Before you start — job brief

A hydrostatic level transmitter does not measure level directly — it measures the head pressure the liquid puts on a tap, and head equals level times density. So the transmitter only reads true if it has been told the product's specific gravity (SG). Get the SG wrong and every reading is scaled off; lose a wet-leg fill or plug an impulse tap and the reading lies in its own way. Never trust the display on its own — judge it against a dip or a sight glass.

  • Hydrostatic level reads head pressure, and head = level × density × 12 (inH2O) — the tx must be told the product SG
  • A wrong SG just scales the reading: a denser product reads high, a lighter product reads low
  • On a closed tank the low side uses a wet (reference) leg — lose its fill and the reading climbs; plug a tap and it freezes
  • Read four things: the indicated level vs a dip, the head the tx senses, the legs/taps, and the SG & zero in config

A level transmitter measures pressure, not level

A hydrostatic level transmitter never actually sees the level. It sees the head pressure the liquid puts on a tap near the bottom of the tank, and it works backward to a level using one fact: head equals level times density. In field units, head (inH2O) = level (ft) × SG × 12, where SG is the product's specific gravity. Twelve feet of plain water (SG 1.00) presses with 144 inH2O; the transmitter divides that pressure back by the SG it was told to get feet, then scales feet to a percentage of the tank height. Every reading on the display is that calculation — which means it is only as trustworthy as the SG, the impulse connections, and the zero that went into it.

The SG just scales the reading

Because level is head divided by density, telling the transmitter the wrong density scales the whole reading by the ratio of the two. Configure it for water at 1.00 and then run a product at 1.20 and the same twelve feet presses with 172.8 inH2O instead of 144 — the transmitter divides by 1.00, reports 14.4 ft, and the display reads 72% against a true 60%. Run a hot, light product at 0.85 and the same level presses with only 122.4 inH2O; the display reads 51%. A denser product reads high, a lighter product reads low, and the error is proportional — it grows with level and vanishes at the bottom of the tank. Whenever a process swaps batches or runs hot, the SG in the transmitter has to follow the product, or every reading is quietly off.

Field note — two pairs that look identical until you check the right thing

Two faults read high: a product denser than the configured SG, and a lost wet-leg fill. They look the same on the display — so check the Legs and the Config. Wet leg full and SG wrong (1.00 configured, 1.20 actual)? It's density — fix the SG. Wet leg empty and SG correct? The reference fill is gone — refill the leg. Two other faults read low: a plugged high-side tap and a lighter-than-configured product. Here the tell is whether the reading tracks. A plugged tap freezes the reading — the head is stuck and won't follow the tank as it fills. A light product still tracks, just scaled — the head is a real, proportional 122.4 inH2O. Same symptom, opposite fix; the distinguishing read is never the level itself.

Wet legs and plugged taps — the impulse faults

On an open tank the low side of the transmitter just vents to atmosphere, but on a closed or pressurized tank the low side has to reference the vapor space above the liquid, and that is done with a wet leg: an impulse line filled with a reference liquid whose column is subtracted from the process head. Keep that leg full and the math works. Let it leak down or boil off and the subtraction disappears, so the transmitter sees a larger differential than it should and the reading climbs toward full — a classic closed-tank fault that always drives the indication high. The opposite failure is a plugged impulse tap: block the high-side line and the sensed pressure can no longer change, so the reading freezes wherever it was when the line plugged and stops tracking the tank entirely. A reading that won't move while a known level changes is a plugged line until proven otherwise.

Zero elevation — the mounting offset

Transmitters are rarely mounted exactly at the tap. When the instrument sits below the tank's bottom connection, the liquid standing in the impulse line adds a fixed head that the transmitter reads as extra level; mount it above and the offset goes the other way. The calibration step that cancels this is the zero elevation or suppression setting, and skipping it leaves a constant offset on every reading — a true 60% might display as 85%. The signature is exactly that constancy: because the error does not scale with level, an empty tank still reads high (about 25% in that example). An offset that survives to an empty tank is a zero problem, not an SG or span problem, and it is fixed in the zero, not by re-ranging.

The discipline

Four reads name every fault here. Compare the indicated level to a dip or a sight glass — that is the ground truth, and a disagreement is your first flag. Read the head the transmitter senses: is it a real, proportional pressure, or stuck? Check the legs and taps for a lost wet-leg fill or a plugged line. And verify the SG and the zero in the configuration against the actual product and the actual mounting. A high reading is a denser product, a lost wet leg, or an unset zero; a low reading is a lighter product or a plugged tap — and the three reads you did not lead with are what tell them apart.