Changing vapour phase and its effect on radiation-based level measurement

Changing vapour phase conditions inside a process vessel can have a significant impact on the accuracy of radiation-based continuous level measurement.

Radiation-based continuous level gauges provide highly reliable measurements because no part of the instrument system is exposed to the process fluid and requires penetration into the process vessel. However, to ensure accurate and repeatable measurement, users should carefully consider the operating conditions inside the vessel. A dynamic vapour space, where the atmosphere above the process fluid changes due to the reaction of the process fluids or changes in temperature and pressure, can have a significant effect on the accuracy of the measurement.

Calibration of any radiation-based measurement system requires referencing the system at known process conditions. For continuous level applications, this means referencing the system at known process levels within the measurement span. The best practice procedure is to acquire references at 10% increments of the measurement span at operating conditions. Often this procedure is not achievable due to a limited ability to alter the process level during operation.

What are often overlooked are the effects the vapour space has on the performance of the system. If, for example, the system is calibrated under conditions that do not produce the same vapour conditions (specifically vapour density) as are present when the vessel is operating, one can then expect the system to have inherent systematic error.

A specific example of such an operating process would be ammonia storage tanks. Liquid ammonia storage utilises pressure or refrigeration systems to maintain the liquid phase. If the process fluid temperature rises due to solar heating, the liquid ammonia can vaporise and will increase the density of the vapour space above the liquid. The increase in vapour density will impact the transmission of radiation through this volume above the process fluid and cause an error in the radiation-based level measurement. The same phenomenon can be realised through a decrease in pressure in the vessel.

Another example of changing pressure influencing the output from a radiation-based level measurement system is observed in a hydrofluoric acid storage drum. One method of transferring the process from the storage drum to the process stream is done by differential pressure between the two vessels. In order for the process fluid to be moved from the storage drum, the drum must be pressured to a level greater than the operating pressure of the working vessels, and the converse is required to transfer from the working vessel to the storage drum. During the pressure changes the vapour phase is altered, causing the vapour density to increase or decrease depending on what action is required. The error in measured level can present issues when accounting for inventory of acid in the storage drum and can present issues when loading the storage vessel from truck delivery of the hydrofluoric acid.

Consider the following application, shown in Figure 1: the desired vertical measurement span is 3658 mm and utilises a single detector and a single 45° collimated source. The vessel has a 3658-mm inside diameter and 13 mm of steel wall thickness, plus 50 mm of insulating material.

Figure 1: Example vessel.

In the chart of Figure 2, the solid line curves represent the response curves of a continuous level for this measurement. The dashed line curves, right-hand scale, represent the error in the measurement when referenced to the empty vessel condition. The variations of the curves are due to the addition of saturated steam at various operating pressures. The specific vapour density values shown are 0.6, 1.1, 2, 3.66 and 5.15 kg/m³. The red curve is shown as a reference, with 0 kg/m³ upper phase density.

It can be seen from the graph that as the vapour space density increases, the signal measured by the detector decreases in an exponential fashion. When compared to a system referenced at a zero upper phase density, the errors can become substantial in terms of the measurement span. It is also observed that the potential error decreases as the process value increases. This is due to the fact that as the process level rises, the vapour space represents a smaller proportion of the measured volume. Therefore, the linearisation curve approaches an approximation of a low upper phase density.

Figure 2: Response curves under vapour phase variation. For a larger image, click here.

It should be clearly stated that the true errors are dependent upon several variables: expected upper phase density and variations from the expectation; measurement span/vessel geometry; the number of sources used for the measurement; and the expected operating point. All of these factors should be considered when specifying equipment for a continuous level application.

A solution for the above conditions can be employed by performing a direct and continuous measurement upon the vapour space. With the ability to monitor the changes in vapour space density, the measurement signal and linearisation of the detector output can be compensated to reduce the inherent error discussed above. Essentially, this vapour space measurement provides a reference that the continuous level measurement can be compensated against.

The reference can be accomplished with a separate radiation-based detector to provide a reference for the continuous level measurement. It employs the same radiation source as the continuous level measurement and therefore utilises the same transmission path through the vessel contents as the continuous level detector. The reference should be made through a radiometric measurement because this technology is the only one that portrays vapour density fluctuation. If one would attempt to measure the vapour space via pressure only, the density could still vary based upon differences in temperature or vapour space composition.

Figure 3: Process value over time showing vapour phase compensation. For a larger image, click here.

Figure 3 shows an application of two detectors summed together (top level and bottom level) to provide a continuous level measurement output shown by the red trend line (overall level). The vapour compensation output is shown in pink, which is the result of a direct measurement of the vapour space above the process fluid. Near point 0.88 on the horizontal scale, a change occurs in the process vessel. The operating pressure is reduced and the vessel is being emptied. If vapour compensation was not provided the level indication would have fallen well below the actual process level, and the indicated level would show less material in the vessel than was truly present.

In Figure 4, trend data taken from the instrument during a depressurisation and de-inventory cycle is shown. The blue trend line indicates a vapour-compensated level indication and the red trend line shows the same instrument output without the vapour compensation function. Both trends are representative of the total level output from the measurement system. Before depressurisation (point 20 on the horizontal scale) there is relative agreement between the trend lines, showing that the system was calibrated under pressurised operating condition. However, after depressurisation and de-inventory of the vessel there is substantial variation, approximately 30% of the indicated span, between the two curves. The consequences to the system operators would be the belief that the vessel has been reduced to 20% of the measurement span, when in reality the true level was closer to 50%.

Figure 4: Process level trend showing vapour phase compensation. For a larger image, click here.

There is an additional advantage that vapour density compensation offers, and it is related to the ability to calibrate the measurement system. Since it is often difficult to change the process level while in operation, and calibration of the radiation-based measurement requires several reference points throughout the measurement span for proper calibration, there is a conflict of needs. When vapour density compensation is added to a measurement system, calibration can occur at non-operating conditions and will compensate the measurement for the increased vapour space density observed when in typical operation. In turn, an error is minimised when brought to operating pressure/vapour density conditions. This aspect has the effect of simplifying the calibration process and balancing operation limitations with calibration requirements.

In summary, we have discussed the effects of a dynamic vapour space and the negative impacts that are possible on the repeatability and accuracy on a radiation-based measurement. There is a tested and proven solution for accommodating these operational issues. This solution can actively compensate for the errors caused and can simplify the calibration process by referencing the radiation signal. The solution can be employed when conditions are known to have a dynamic vapour space due to temperature or pressure fluctuations or variations in vapour material composition, as well as when operational conditions do not allow for the optimal calibration conditions.