Average tank liquid temperature is a single representative temperature for the entire body of product in a tank, built from measurements taken at several depths rather than from one convenient spot. It matters because a large tank does not sit at one uniform temperature: warm product floats above cooler product, sun heats the top, and the ground cools the bottom, so the liquid is stratified. Since temperature drives volume correction, using a single spot reading on a stratified tank corrects the whole inventory as if it were all at that one temperature, which it is not. Averaging across the depth is how the correction is made to reflect the tank as a whole.
Average tank liquid temperature in one line: Average tank liquid temperature is a representative temperature for the entire liquid column, computed from measurements taken at multiple depths instead of one spot. It is needed because tanks stratify, and temperature correction of custody-grade volume must reflect the whole product mass, so a single spot reading on a large or layered tank is not accurate enough.
Temperature is the input that converts a measured volume into a corrected, standard-condition volume, so its accuracy flows straight into the reported barrels. On a small, well-mixed tank, the liquid is close to uniform and a single reading is a fair stand-in for the whole. On a large tank, that assumption breaks down. Product delivered warm and left to sit develops layers; solar gain warms the upper portion while contact with the ground cools the lower portion, and the difference between top and bottom can be significant.
In that stratified condition, a spot temperature only describes the layer where the sensor happens to sit. Correct the entire inventory using a reading taken near the top and you overstate how warm the cooler lower product is; use a bottom reading and you understate the warmer upper product. Either way the temperature correction is applied to the whole volume as though it were all at one temperature, and the resulting standard volume is biased by however far the spot reading sits from the true average of the column.
The severity scales with the tank and the conditions. A tall tank holding recently received warm product on a sunny day can carry a real vertical temperature gradient, and the larger the volume, the more each fraction of a degree of averaging error is worth in corrected barrels. This is precisely the situation custody transfer cannot tolerate, because the whole point is a defensible number, not a plausible one that happens to be a little off.
The remedy is to sample temperature at multiple depths and combine the readings into an average that represents the column. Manually, this has traditionally meant lowering a portable thermometer to several defined levels, top, middle, and bottom portions of the liquid, letting it equilibrate at each, and recording the readings. The individual spot temperatures are then combined, ideally weighted by how much volume each represents, to produce an average liquid temperature rather than a single point.
Permanently instrumented tanks use a multi-point averaging temperature element, a single probe carrying multiple sensors spaced along its length. As the sensors that are submerged report and those above the liquid are excluded, the device produces a temperature that already reflects the wetted depth, updating as the level changes. This removes the labor and inconsistency of manual multi-level dips and gives a continuous average instead of a snapshot taken whenever someone climbs the tank.
The word average is doing real work here. A simple mean of top, middle, and bottom is only correct if each represents an equal slice of volume, which is rarely exactly true, especially as level changes on a tank with non-uniform geometry near the bottom. A volume-weighted average, which weights each temperature by the portion of the inventory it represents, is the more accurate approach, and it is exactly the kind of continuous calculation that is tedious by hand but trivial for an instrument or a computer.
A modern automatic tank gauge with a multi-point temperature element does the averaging continuously and feeds the result directly into the net volume calculation. It knows the live level, so it knows which sensors are submerged, and it can weight the contributing temperatures by the volume each represents to produce a volume-weighted average liquid temperature in real time. That value, rather than a lone spot reading, is what drives the temperature correction, so the correction reflects the whole product mass at every moment.
In a SCADA context, this turns temperature from a once-a-shift manual chore into a live, historized signal. Merobix can record the average liquid temperature continuously alongside level and the corrected volume, so the temperature that drove any given custody figure is preserved and auditable rather than reconstructed from a logbook. Continuous averaging also means the correction is right during the day and at night, on a full tank and a low one, without waiting for someone to go take multi-level readings.
Historizing the average temperature also exposes problems that a single manual reading would mask. A failed or fouled sensor on the averaging element, or a genuine strong stratification event after a warm receipt, shows up in the trend, letting operators respond rather than unknowingly correcting a whole inventory with a compromised temperature. Because the average is what net volume depends on, keeping it live, weighted, and recorded is a direct contribution to the accuracy and defensibility of the custody numbers the SCADA system reports.
Large tanks stratify, with warmer product floating above cooler product because of solar gain on top and ground cooling at the bottom. A single spot reading only describes the layer where the sensor sits, so correcting the whole inventory with it biases the standard volume by however far that spot is from the true average. Custody-grade net volume needs a temperature that represents the entire liquid column.
It is a single probe carrying several temperature sensors spaced along its length inside the tank. The sensors that are submerged report while those above the liquid are excluded, so the device produces an average temperature that already reflects the wetted depth and updates as level changes. It replaces the labor and inconsistency of manually lowering a thermometer to several levels.
It is an average of the multiple depth temperatures in which each reading is weighted by the portion of the tank's volume it represents, rather than a simple equal-weight mean. This matters because equal slices of height do not always hold equal volume, especially near a tank's bottom geometry. A modern ATG computes this weighted average continuously, which is tedious by hand but straightforward for an instrument.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.