What Is a Densitometer?

Back to Blog

How a Densitometer Works

The most common inline densitometers use a vibrating element - a tube, tuning fork, or vibrating cylinder - kept oscillating at its natural resonant frequency. The fluid to be measured fills or surrounds that element. The resonant frequency of a vibrating body depends on its total mass, so as the fluid's density changes, the mass moving with the element changes, and its resonant frequency shifts. The transmitter measures that frequency precisely and converts it to density, applying temperature and often pressure compensation because both affect the reading.

Because the measurement is based on frequency, it is very stable and repeatable and does not depend on flow rate. Some measurements come from a dedicated densitometer in a sample slipstream; in many modern installations, a Coriolis meter provides density as a native output because its vibrating tubes measure density in exactly the same way while also measuring mass flow.

Why Density Matters and Oil and Gas Fit

Density ties volume and mass together, so a live density value is essential for accurate measurement. It converts a mass flow to a volume or a volume to a mass, corrects observed volumes to standard conditions, and yields API gravity - the standard way crude and products are graded and priced. In custody transfer, an inaccurate density directly distorts the quantity and quality that money changes hands over, which is why real-time densitometry sits alongside the flow meter on a custody skid.

In oil and gas, densitometers measure crude, refined products, NGLs, and produced fluids for custody metering, API gravity, blending, and water-cut context. The transmitter reports density (and often computed API gravity) to a flow computer, which combines it with flow and temperature to compute net volumes and mass. A cloud SCADA such as Merobix reads those density and gravity tags from the flow computer over Modbus or DNP3, so density trends and out-of-spec conditions are visible and alarmable remotely.

Installation Details That Protect the Reading

A densitometer only reports the density of the fluid actually touching its element, so installation is mostly about making that fluid representative of the stream. On a slipstream installation the sample loop must flow continuously - a dead or lazy loop measures old, cooled fluid rather than the process. The takeoff should come from a point where the fluid is well mixed, and the loop should be short and insulated so the sample reaches the element at essentially process temperature, because every degree of difference is an error the compensation has to work to remove. Orientation and mounting follow the manufacturer's datasheet, since drainage and gas-shedding behavior differ between tuning-fork, cylinder, and tube designs.

For liquid service, entrained gas is the dominant enemy. Even a small fraction of bubbles drops the bulk density and disturbs the oscillation, so the meter should sit where operating pressure keeps gas in solution - not just downstream of a control valve that has flashed the fluid. Where the process itself can carry gas, a conditioning or deaeration arrangement ahead of the element is a design decision to settle early, not a retrofit after the readings disappoint.

Verification Against an Independent Reference

Custody applications do not take the densitometer on faith; they check it against an independent reference on a schedule set by the measurement contract. The classic method draws a sample at the densitometer while logging its live reading, determines the sample's density by pycnometer or in the lab, corrects both to the same temperature and pressure basis, and compares the two. The mechanics and the acceptance logic are covered in the densitometer verification reference.

The offset between meter and reference is worth trending even when every individual check passes. A stable small offset is a characteristic of the installation; an offset that steps or begins to drift is a message - coating, drift, or a sampling problem - arriving before the meter actually fails a check. The verification interval itself is site- and contract-specific, but the discipline of logging and trending each result costs little and catches the slow failures that a pass-fail record hides.

Failure Modes and What They Look Like

Symptom	Likely cause
Reading drifts or steps high	Wax, scale, or coating adding mass to the vibrating element
Noisy reading biased low	Entrained gas or flashing in the sample
Offset that follows process temperature swings	Temperature compensation fault or poor loop insulation
Reading frozen while the process moves	No flow through the slipstream, or a drive or electronics fault
Slow divergence from lab samples	Calibration drift or a non-representative sample point

The physics explains the directions. Coating adds moving mass to the element, which lowers its resonant frequency, and a lower frequency reads as higher density - so fouling biases high. Gas does the opposite: it lowers the true bulk density and disturbs the oscillation, so gassy readings sit low and jumpy. Those signatures let an operator separate a process change from an instrument problem at the trend screen, before anyone opens the sample loop. A useful cross-check is whether the flow computer mass and volume figures still reconcile, because a density bias shows up there as a growing imbalance between the two.

Frequently Asked Questions