Sensor tips: Making the most of multivariable sensors
June 28, 2011 by Ian Verhappen
In the past, I have talked about how smart instruments are able to continuously monitor their own health and report diagnostic information about their status. In addition, bus-based and smart instruments can be configured to use one of several measurements as their process variable (PV) or output signal. In some cases, the choice of measurement performs additional manipulation – e.g., the square root for flow or the single “raw signal,” such as the capacitance from a differential pressure cell. In other cases, the microprocessor-based device calculates an output based on inputs from multiple sensors – typically temperature compensation of the basic measurement.
Despite the fact that we have been performing basic calculations with our field devices for many years, very few people are willing to take that next step and use a secondary sensor reading from the same device, even for a monitoring signal – keeping the analog output in a HART device as the PV. As we know, bus-based transmitters can send multiple variables from a device by configuring the appropriate function block with no additional impact on infrastructure other than the required bandwidth to transmit the data.
The most easily identifiable multivariable transmitter application is for measuring flow. The most obvious candidate for multivariable signals is a Coriolis flowmeter, since these devices are capable of measuring mass, flow, density and, depending on the configuration, other calculated variables as well.
The simplest multivariable flow transmitter is the differential pressure (DP) cell, which is normally used for orifice flow – the most widely used flow measurement technique. Orifice flow can also provide bulk pressure measurement of the upstream or downstream pressure tap. Therefore, it is possible to use a single transmitter to provide both pressure-compensated flow and bulk stream or process pressure at no additional cost.
Another commonly used flowmeter is the vortex meter. For several years now, multiple vortex meter manufacturers have been adding a temperature sensor and associated temperature compensation calculation for the flow calculation into their meters, especially those with digital communications capability. Because volume flow, at least for custody transfer, is always at standard conditions – an agreed upon pressure temperature – integrated temperature correction is beneficial. For incompressible fluids (liquid) where the contribution of pressure to the correction factor is negligible, having a temperature-compensated measurement is often sufficient. In most cases where the stream composition does not change much, temperature compensation allows for calculation of mass flow in the device. Mass flow is the preferred method of flow measurement because it is independent of flow conditions.
All of this is fine for liquid flows, but there are a lot of vapour streams that need to be measured and controlled. Vortex meters are often used for vapour flows; however, because gases are more affected by changes in pressure, temperature compensation alone is not sufficient to calculate the mass flow. The Ideal Gas Law (n = PV with n = m which rearranges to m = (MW) PV ) allows for the direct calculation of mass of a gas stream with constant composition from volume flow, pressure and temperature measurements.
In the above equations: m = mass; MW = molecular weight of gas; n = number of moles of gas; P = pressure; V = volume; T = temperature; and r = “universal” gas constant, which is a function of the units for P, V and T.
Until recently, calculating gas mass flow required multiple sensors; however, there are now two manufacturers – Krohne and Honeywell – who have integrated pressure- and temperature-compensated vortex meters, so users can calculate mass flow for vapours/gases with a single meter. Several other manufacturers also have this calculation capability, but require an external pressure signal to be connected to the vortex meter.
The preferred way to measure steam flow is in kg/hr and, fortunately, because the composition of steam does not change, the calculation of steam flow with a temperature- and pressure-compensated vortex flowmeter is relatively straightforward.
Using a single sensor to provide multiple measurements comes with the risk that failure of a single device could have a larger impact than in the past; however, this risk is balanced by several other factors, including the fact that these smart transmitters also come with diagnostics to predict their health and status, so that proactive maintenance can be performed before they fail.
There are many other savings beyond the cost of additional sensors, such as:
• Process connection – a typical nozzle is normally several thousand dollars by the time it is installed;
• Cable – a single cable must be run from the sensor to the field junction box and then an additional pair added to the multi-conductor cable to the control system; and
• I/O card – each signal requires a port on the host I/O card to allow the measurement to be captured by the control system.
In addition, when you buy additional sensors, there are the associated expenses for the original design, spare parts and maintenance, which can add up to significant costs.
It continues to surprise me that more people do not use multivariable sensors. Hopefully awareness will lead to activity and a new trend can begin.
Ian Verhappen, P.Eng. is an ISA Fellow, ISA Certified Automation Professional, and a recognized authority on Foundation Fieldbus and industrial communications technologies. Verhappen leads global consultancy Industrial Automation Networks Inc., specializing in field level industrial communications, process analytics and heavy oil / oil sands automation. Feedback is always welcome via e-mail at firstname.lastname@example.org.