Pressure gage ranging and accuracy
Dear Control Engineering: I was readingand I had a question coming out of that discussion. The author made the following statement: “The pressure in the line may be upwards of 1,000 psi, but the pressure drop is usually very small. Using two separate gage pressure devices and subtracting the readings would result in the combined error of two devices having wide measuring spans. The result would likely be a very poor measurement seldom suitable for control.” Could you explain that in greater detail?
That statement was made in the context of using a DP (differential pressure) sensor to measure flow using an orifice plate flowmeter. His point is that a single DP device is better than trying to do the same job with two pressure gages and calculating the difference.
Here’s why: Using his example, let’s say your process line is running with a pressure of 1,000 psi, and you have an orifice plate. Under normal operating conditions, the pressure drop across the plate is 5 psi. That means if everything is perfect, the upstream gage reads 1,000 psi and the downstream gage reads 995 psi. For appropriate ranging, both gages are identical with a scale of 0-1,500 psi. The problem emerges when you consider the accuracy of the two devices. If they have an accuracy rating of ±1% full scale, that means any reading can be high or low by 1% of the full scale, or 15 psi. So the first gage could indicate 1,000 psi as anything from 985 to 1,015 and still be in tolerance. Similarly, the second gage could indicate 995 psi as anything from 980 to 1,010 and still be in tolerance. Do you see the problem? You’ll never be able to know when your pressure drop is, in fact, 5 psi.
The differential pressure device, if it has a range of 0 to 15 psi, will give you a much higher accuracy since 1% of 15 is 0.15 psi.
Now, the tricky part is making sure you’re gage isn’t exposed to 1,000 psi on one side and 0 on the other. This could damage it, so you will have to ensure that the device won’t see pressure level imbalances outside of its ability to withstand them.
Peter Welander, email@example.com