The special requirements of clean-in-place procedures can introduce inaccuracies in temperature and pressure readings.
Clean-in-place procedures require the site cleaning of all the internal components of a process system, including the pressure sensor head, after each batch run to prevent bacterial growth and contamination, so as to maintain the highest standard of product quality. This practice is common to a variety of industries, such as food and beverage, pharmaceuticals, fats and oils, paint and ink, cosmetics and perfumes, prepared animal feed, chemical manufacturing and clean gas systems.
Clean-in-place units must survive high temperatures and cope with rapid temperature changes that can impair pressure sensor accuracy. This specification is generally defined as the sum of the pressure sensor's linearity, hysteresis and repeatability.
A unit's accuracy is affected by temperature effect, which is seen as an off set of the zero and span of the sensor. Both are usually expressed as a +/- percentage change in the rated output per °F change in ambient temperature over the compensated temperature range of the particular model. This could, for example, be expressed as 0.0025/ °F change on both zero and span.
The process engineer uses these temperature effect numbers to calculate the applied accuracy of a given pressure sensor, taking into account the actual temperature of the application.
Temperature-induced expansion is a problem for all sensors and instruments. Pressure sensors are typically constructed of various types of metal, welded together to form a unit. To achieve optimal thermal stability, these metals must be matched in terms of their thermal expansion coefficients and high-tech welding processes must be used to join them. When extra large metal components (e.g., a sanitary pressure flange) are part of the pressure sensing unit, the thermal expansion problem becomes considerably more complicated.
The temperature gradient that develops across a pressure transducer body defies the compensating techniques used by most pressure sensor manufacturers. The reason is that the temperature compensation is electrical and located electrical and located at the pressure sensor's back end with the rest of the electronic components. Therefore, while the front end of a pressure sensor may see temperatures of 200°F - 400°F, the electronic compensation components are in a very different climate (see Figure 1).
As the process medium comes into contact with the pressure sensor diaphragm, the pressure sensor experiences a temperature gradient that heats first the coupling assembly and then the pressure sensor to different levels at different locations.
This temperature gradient effect also influences response to temperature change, a particularly thorny design problem. If the nature of the application permits the unit to stabilize at some temperature before a reading is taken, the pressure transducer's accuracy rating is reliable. In the real world, however, most pressure transducers take from 1 to 2½ hour to stabilize at a new temperature when they move from ambient to above 200 °F, and few processes are slow enough to allow that amount of time for stabilization.
Pressure sensor manufacturers have solved this problem in two ways. Smart pressure transmitters were developed with on-board electronic temperature compensation, which is one of the factors that make them much more expensive than the conventional variety. A fairly recent innovation combines basic pressure sensor technology and a clever mechanical design (see Figure 2) that permits non-smart pressure sensors to track temperature changes as fast as 200°/ minute, which is more than adequate for many processes.
These devices are useful when it is necessary to precisely control some time-related part of the process, where the temperatures are changing, e.g., adding ingredients at the proper time and/or in the proper quantity or opening and closing valves or doors.
An Ultra Series pressure sensor with a 1 ½ inch, clean-in-place (CIP) flange was electronically given an intentional zero offset to simulate a constant pressure. As the temperature of the process medium was increased, the average temperature of the pressure sensor rose, causing a shift in the simulated pressure reading. The observed change, called zero offset, illustrates temperature-induced pressure sensor inaccuracies. Approximately 50 minutes into this test, the pressure sensor was removed from the ClP clamp and immediately immersed in room temperature water. A sharp decline in both the process medium and the pressure sensor temperature readings, and a concurrent decline in the zero offset line, show the pressure sensor's almost instantaneous response to the extremely rapid temperature change. Most pressure sensors would exhibit the same response, but over a 15-30 minute time period instead of only a few minutes.
The term sanitary (hygienic in Europe) denotes a portion of the CIP market that makes products for personal use or ingestion by human beings. The Sanitary Standards Council, in conjunction with the FDA, has developed standards that must be met for a manufacturer to display the 3A label. These standards pertain to both the mechanical design and the fluids that transmit pressure inside the sensor.
Diaphragms (see Photo 1), for example, must fit flush and have no crevices that could capture and retain impurities.
Photo 1. The flush-mounted, smooth-surfaced diaphragm of this 3A-rated pressure sensor is sensitive enough to provide a timely response to changes in temperature and pressure with 0.15% F.S. accuracy. Although massive, the diaphragm is sensitive enough to measure pressure from 600 psi down to 3.5 psi. The sensor assembly withstands the 400E° F temperatures associated with ClP procedures.
Temperature compensation is also complicated by the 3A guideline dictating the use of an edible fluid to transmit pressure sensed on the flush diaphragm of the sensing element. This fluid is somewhat compressible, and has a thermal expansion coefficient quite different from that of any of the metals used in pressure sensor construction. Only within the past few years has it become possible to substantially minimize the impact of the edible oil.
Furthermore, if the pressure sensor manufacturer does not provide the proper controls, these fluids can fill up with air molecules that separate from the fluid at process temperatures. As bubbles are formed, the gas expands or is compressed as space allows, resulting in erroneous pressure readings.
The edible pressure transmission oil introduces yet another variable, called the orientation effect. Each oil has a given density (weight), which the sensor will detect as a pressure. The amount of pressure exerted by the oil depends on how the pressure sensor is installed in the process piping. If a unit is installed with the diaphragm up, for example, the total weight of the oil is on the pressure sensor; if the unit is positioned with the sensing head down, it sees no oil weight. Different degrees of angle in a horizontal mounting will produce varying weights. The user will see this phenomenon as a zero shift.
So, as vibration or system maintenance changes the pressure sensor’s orientation, a skilled technician or an engineer will have to go to the sensor site, which is often remote and in hard-to-access spaces, and adjust a potentiometer on the unit to eliminate the zero shift. The degree of orientation effect is determined by sensor design, and it varies from one manufacturer to another. Recent advances have made this effect almost negligible, with the availability of units with 0.03 psi/g maximum error due to orientation. There are few processes that would be bothered by this small variation.