This column is moderated by Béla Lipták (http://belaliptakpe.com/), automation and safety consultant, former chief instrument engineer of C&R, and former Yale University professor of process control, who is also the editor of the Instrument Engineer's Handbook. If you have automation-related questions for this column, write to email@example.com.
Q: I need your help in installation of a pressure transmitter for dusty service. We're working on blast furnace stack pressure measurement. It's full of dusty particles. We suggest a scheme, attached herewith, wherein we are purging the impulse tube constantly with low-pressure nitrogen and also blasting nitrogen at regular intervals using solenoid valves (Figure 1). Please suggest any alternate option or your opinion on the same.
Pressure is near 3 bar, and temperature is around 1200 ºC. I have attached a scheme for installation of the pressure transmitter, wherein we propose to purge the impulse line with low-pressure nitrogen constantly so as to avoid choking of the impulse lines. Also we have made provision for blasting at regular intervals using a solenoid valve that is operated by our PLC system.
A: As shown in Figure 2 below, I would simplify the configuration. I would make the furnace nozzle larger and install it with a 45° slope. I would also use a seal diaphragm on the pressure transmitter and make it an indicating one, so that the operator can see the pressure without going to the control room. If the nitrogen blast pressure is higher than the normal purge pressure, I would place a check valve in the purge line, and instead of a rotameter, I would use just a needle valve. This will reduce maintenance.
Chemical seals are temperature sensitive—I suspect the blast furnace stack is likely to be too hot. There are pneumatic pressure repeaters but these will become coated—I assume from your design that the dust is adherent, which is why the gas blast. A small spring return pneumatic cylinder driving a ramrod scraper through the process tapping may be worth considering as part of the intermittent blast (use the nitrogen to drive the cylinder and the exhaust to flush as it retracts). A:
I once used a NaK alloy chem seal fill for a bitumen blowing column level transmitter, but a technician took the easy way out to undo the seal from the vessel. The NaK ignited his boots. I have never tried it again. Too many folk don't care to read warning notices.
Ian H. Gibson
I agree—using a process seal instead of an impulse tube is likely to work better, unless the pressure differential is so small that the added mass of the seal changes the sensitivity of the instrument. Also, using a process seal doesn't completely evade the problem of particulate adhering to the seal itself, which will also degrade the measurement. Fly ash is ugly, nasty, abrasive, corrosive and adheres to steel really well, as I found out when I was putting in nuclear point level gauges for fly ash level detection back in the 1980s. In that case, the use of nitrogen purging is the only way I know of to keep the impulse tubing clean and unplugged. It may be a pain, but it works.
A: For light, dusty service, purging the impulse line (0.25 in. O.D.?) may work. For heavy, dusty service, purging even every hour may not work, and can damage the sensor in the pressure transmitter. You may need to change the small impulse line into a 3-in. pipe (or bigger) with a flange, and then use a pressure transmitter that uses a flanged pressure diaphragm for sensing. One supplier (Rosemount) has them. Sometimes even the 3-in. pipe can get plugged up, in which case, you will have to engineer a purging system (a 3-valve "T" system) for purging the dust. Make sure that the pressure diaphragm is isolated before purging. Do not use gas containing oxygen (or any combustibles) for purging dust particles to avoid explosion.
A possible grey area between mechanical discipline and instrument discipline could exist when a pressure sensing line (impulse lines), typically 0.25 in. O.D., is changed into a 3-in. sensing pipe. In many process plants in North America, the first valves that connect to any process are the responsibility of the mechanical engineers. In the case of a 0.25-in. pressure sensing line, a threadolet and a mechanical isolating valve is normally used for connecting the line to typically a 3-valve instrument manifold on which a pressure transmitter is mounted by instrument discipline.
In the case of a 3-in. pressure sensing pipe connecting to a process, a double-block-and-bleed valve is often engineered; this provides an isolation of the instrument system when maintenance of the pressure sensor (a 3-in. pressure diaphragm with flange) is required. You may like to change your ball valve into a double-block-and-bleed valve if a 3-in. pressure sensing pipe is implemented. When a single ball valve was used in existing practice, you may need to consider the need to protect your pressure sensor when the ball valve is turned on suddenly. For processes with toxic gases, the bleed valve (in the double-block-and-bleed) is often connected by a pipe vented to a safe area defined by area classification. Operation procedures are often written on how the double-block-and-bleed valve is to be operated during start-up and shutdown of the plant. Although the purging process by instrument discipline looks simple, a hot permit for purging will be inevitable when the plant is running.
The use of purging rotameters in very dusty applications is not recommended owing to high maintenance. I can see that it is likely to be used during manual purging of the impulse line. For automatic purging (by energizing your solenoid valve by PLC), you need to modify your existing purging scheme by adding more solenoid valves because of the need for automatic isolation of the pressure sensor before purging and of the manual purging system. This suggests that your PLC ladder logic will likely need to be modified.
A: The proposal as outlined might work well. Preferably the impulse lines should be relatively large and arranged to let gravity be your friend and let the dust flow down and back out. I suspect that the pressures involved might affect the listed accuracy of the pressure transmitter. The loss of accuracy might be less of a problem than the problems of plugging. Moore Products (now Siemens) once made a sensitive pneumatic pressure repeater with a large diaphragm as the seal. Any gas purge should have the flow rate set by a needle valve with a relatively high pressure to that valve. This keeps the flow rate constant.
A: This scheme will work; however it has too many fittings, and dependencies like rotameter adjustments will make it a maintenance headache. A sealed diaphragm type transmitter mounted directly on spool piece after the first flange on the pipe with a N2 purge connection on the spool piece will help. This can be manually operated too if you wish to save some money
H. S. Gambhir