Solving Cold-Related Problems with Pressure Instruments

  • February 15, 2016
  • Emerson Automation Solutions
  • Feature

Field devices, especially pressure measurement installations, can experience maintenance problems and erratic performance in cold weather. Effective product design, installation and maintenance procedures can minimize these issues.

By Mark Menezes, PEng., Canada Manager - Emerson Process Management, Rosemount Measurement

Many process plant operations are located in areas where seasonality can be an operational issue. Hot summers alternating with long winters bring severe temperature extremes which can affect equipment of all sorts, particularly the sensors and actuators monitoring process units. Cars are harder to start when temperatures fall below -15 ºC (5 ºF), and valves can stick and field instruments can fail or lose accuracy.

Why Temperature Matters

For purposes of this discussion, we will explore the winter end of the scale. Low temperatures in this context are those common to higher latitudes and elevations where process plants are located, bottoming out at around -40 ºC (-40 ºF).

Some temperature effects are closely tied to specific numbers. The freezing point of water at 0 ºC (32 ºF), is a critical one given how much water is around us. When water freezes, hydrogen bonding causes it to expand with force able to burst pipes and crack foundations. Many liquids have lower freezing points, but become more viscous as the temperature goes down. This applies to many oil-based substances including lubricants.

Anyone who has tried to start a car in the cold sees many of these symptoms firsthand. The battery loses power and gasoline is more difficult to change from liquid to vapor. Those old enough to remember the days of cars with carburetors know this well, but today’s cars are much easier to start in the cold because they have been designed to be more tolerant of low temperatures. Fuel-injection systems work well even in below freezing conditions. In the same way, many field instruments and actuators available today are designed to work much better in bad weather than past designs.

Electronics in the Cold

Many electrical devices actually work better in cold weather. The electrical resistance of conductors goes down, and this relationship of temperature and resistance makes sensing technologies such as RTDs and thermistors possible. Many devices such as motors depend on their ability to dissipate heat for peak performance, and can thus operate easily in the cold.

However, problems begin to develop when sophisticated devices, such as the A/D converters and other elements of field device transmitters, were designed with more normal temperatures in mind. The electrical characteristics of semiconductors aren’t always the same at low temperatures, and combinations of dissimilar metals within the circuits can create microscopic thermocouples, with results not always predictable or consistent.

In response, electrical designers have built better circuits and the components have improved so the characteristics are better understood and their effects reduced, as shown in Figure 1. Moreover, most sensor technologies require some degree of thermal compensation in all circumstances.

Figure 1: Even when coated with frost and in -50C weather, this pressure instrument continues to function reliably and accurately.

A capacitive or strain-relief pressure instrument has a temperature sensor built into it already, and most vendors have found ways to increase the effective operating range in both directions on the temperature scale. A sophisticated pressure instrument is probably monitoring ambient temperature via its built-in sensor, and might also measure the temperature of the transmitter’s electronics, as both can affect reading accuracy. Whether you realize it or not, there is probably much happening in the device to ensure an accurate, repeatable reading over the widest possible temperature range.

Fluids are the Bigger Problem

Years ago, people in the north used to “winterize” their cars, which meant changing to thinner motor oil, refreshing the antifreeze and making sure the battery could hold a full charge. A sluggish battery trying to crank a cold engine with thick oil was a challenge. Plant owners in areas where the seasons change drastically from summer to winter face similar issues and winterize strategic pieces of equipment. If they don’t, the first serious cold snap can cause some strategic failures. More than one plant has suffered because a fluid froze and broke a pipe.

One of the perpetual problems when the mercury sinks is frozen impulse lines—those small tubes (capillaries) leading from the process penetration point to a pressure instrument, carrying either the process liquid or some other filler material to transmit pressure to the sensor (Figure 2). Those lines allow the sensor and its associated transmitter to be mounted in a location easier to reach than the actual process penetration, or allow one sensor to connect to multiple points some distance apart.

Figure 2: Lines between the process penetration and instrument must be carefully installed and maintained to avoid freezing and other issues.

The pressure instrument might be performing various tasks. It could be measuring the actual process pressure, it could be using a differential measurement to calculate flow, or it might be using pressure to measure the level in a tank. Those impulse lines have to be filled with something, either gas or liquid, and are described as dry legs or wet legs, respectively. If the instrument is on a steam line, the fluid is probably condensate.

Differential pressure flowmeters are commonly used for measuring steam flow. The impulse lines are wet legs because steam condenses in them, filling them with condensate. Maintenance technicians often expect these to be impervious to cold weather because they are connected to the steam line, which transfers heat down the metal tubing, usually stainless steel. They’re also normally insulated, at least to some extent. Still, it’s an unhappy surprise if the first hard freeze disables the instrument and maybe ruptures the lines, all because technicians don’t realize how quickly heat can be dissipated.

Challenges of Steam Lines

Since the process (steam) and environment are at different temperatures, the temperature along the impulse line will change as heat transfers to the environment. Insulation can slow the change but can’t stop it. This complicates design when the process is hot and the ambient temperature varies, as is common in outdoor installations. If the impulse line is too short, not enough heat is dissipated in summer, and the instrument can become overheated and damaged. If the line is too long, too much heat dissipates in winter and it freezes.

Figure 3 shows how a typical insulated sensing line (1/4 in., 316 stainless-steel tubing) can cool by 140 °C (250 to 110 ºC) in 160 mm (250 ºF in 6 in.) when ambient temperature is 0 °C. At a higher ambient temperature, 40 °C for example, heat dissipation is nearly five times slower, so the same 140 °C heat change happens in 800 mm (30 in.). At an extremely cold temperature—for example -40 °C—heat dissipation is twice as fast, so the same change happens in 80 mm (3 in.).

Figure 3: As this graph shows, heat dissipates quickly from stainless steel tubing, but at predictable rates.

Problems with these impulse lines often cause maintenance technicians to replace them with oil-filled capillaries. The fluid product in the tubes has a higher molecular weight (MW) than water so it can operate at the full steam temperature without boiling off. Some silicone-based products have boiling points well beyond 300 ºC (570 ºF). Unfortunately, the colder end of the line can be a problem. Viscosity becomes an issue with these fill fluids at lower temperatures. Table 1 gives examples of common products and their temperature versus viscosity characteristics.

Table 1 -- Common Fill Fluids: Boiling Point and Viscosity at Selected Temperatures5

Fill fluid

Boiling Point °C

Viscosity @ 25°C (cSt)

Viscosity @ 0°C (cSt)

Viscosity @ –25°C (cSt)

Syltherm XLT





Silicone DC200





Silicone DC704





Silicone DC705





When viscosity increases, response time slows down. A 5 m long capillary tube with an internal diameter of 10 mm filled with fluid with viscosity <5 cSt (CentiStokes) slows response time by 1-2 sec. The same system with a fluid viscosity of >150 cSt slows response time by >30 sec. When the fill fluid solidifies, it provides no response at all.

Finding an All-Season Solution

If a plant is in a location where it is either hot or cold year-round, it is relatively simple to design a solution. However, where temperatures can swing from -40 to 38 ºC (-40 to 100 ºF) over the span of a year, the heat dissipation characteristics of impulse lines change drastically with temperature. When using traditional methods, it is very difficult to create a single passive approach capable of avoiding both freezing in winter and overheating in summer.

One common but expensive alternative is adding thermostatically controlled heat tracing on the impulse lines. Usually these systems only add heat during the winter and can avoid overheating in the summer, however they can double or triple the cost of adding a pressure instrument, require energy to operate, and complicate maintenance tasks.

Newer capillary systems are designed to eliminate the need for impulse line heating without slowing response time. As shown in Figure 4, the seal is directly connected to the vessel or pipe containing hot fluid. The design of the seal and its internal copper tubing are optimized to conduct the right amount of heat so the oil remains in a liquid phase with low viscosity for best responsiveness during the winter, but does not conduct so much heat as to damage the transmitter during the summer.


Figure 4a and 4b: This thermally-optimized wireless pressure transmitter, shown in full and cutaway views, replaces impulse lines with a sealed tube.

For very hot processes, or where the instrument must be located a greater than normal distance from the process, a two-oil solution may be needed as shown in Figure 5. High MW oil is used adjacent to the hot process. This oil provides high-temperature stability and remains hot enough to ensure fast response time. Low MW oil, such as Syltherm XLT, is used after the intermediate seal where the oil is cooler and runs through the capillary to the instrument. This oil retains its low viscosity below –50 °C for fast response even in the dead of winter.

Figure 5: Newer configurations use permanently sealed capillary lines filled with multiple oils, each one optimized to perform in a variety of temperatures.

Eliminating the Root Cause

Steam flow measurement applications in cold climates are one of the most common winter-related problems. One way to eliminate the complexities of impulse lines is to eliminate them altogether. Some flowmeter types are designed as native measuring devices (Figure 6) rather than as an adaptation of a differential pressure instrument. There is no doubt the traditional method works, but dealing with condensate-filled wet legs can remain a maintenance headache.

Figure 6a and 6b: Differential pressure measurements are commonly used to measure flow. These units are configured specifically for this purpose, and thus eliminate many of the problems associated with site-designed impulse lines.

Integrated flowmeters are built as a single unit designed around a more effective configuration, placing the pressure sensor and transmitter above the steam line rather than below it. They are generally pre-assembled, configured for this specific service, and shipped after being leak tested and calibrated. The heat dissipation components are designed to operate in warm and cold environments.

Based on this standardized approach, it’s a simple matter to calculate the heat transmission characteristics based on the process temperature and typical ambient temperatures. Extreme cases may require additional considerations, but these are not difficult to handle. Above-the-pipe mounting and single-unit construction makes installation easier and less expensive. Since there is no need for heat tracing in most situations, the total installed cost is also lower.

Self Diagnostics

Smart instruments are able to monitor the operation of the device and send data to the control system via protocols such as HART, Foundation Fieldbus, Profibus PA, EtherNet/IP, WirelessHART or ISA100.11a. These capabilities allow monitoring of temperature and the performance of the transmitter to ensure the device is performing its best. Even when used with traditional mounting approaches, a smart instruments can tell the control system when impulse lines are clogged or frozen. It can also let the control system know the ambient temperature, which can be useful for correcting pressure and other measurements.

Weather-related problems will exist to some extent as long as oil, natural gas and minerals have to be extracted in difficult environments. In some extreme situations, users may be forced to install heated enclosures where all other simpler approaches fail, but these are declining as field devices become better at dealing with difficult environments. The electronic devices in transmitters are less sensitive to temperature, or do a better job of correcting for it. Modern mechanical configurations of sensors, mounting and piping are also less influenced by cold and heat.

Repairing failed devices in the winter is hazardous, expensive and hard on technicians. Fortunately, the need to perform such tasks is becoming far less frequent. Many technologies are available to help eliminate some of the traditional challenges, and utilizing dual oil fill systems with smart instruments can significantly simplify measurements in cold conditions.

Did you enjoy this great article?

Check out our free e-newsletters to read more great articles..