We are doing a startup of a new biotech facility. We are monitoring (but not controlling) differential pressure. This is done by installing a single DP instrument for each room of interest and referencing it back to a common header. Then interelationships can be figured arithmatically in the building management system to determine DPs between rooms. Unfortunately, our header seems to be building some pressure (positive at times and negative at others). It is vented to the outside via a specialty diffuser (which is supposed to eliminate wind effects). I wouldn't care so much about the offset, but it is not uniform. Does anyone have experience with similar systems so we can compare notes. This is my first pass at using a common reference header like this, and until I resolve it the outputs from the DPM's are essentially crap.
Eng-Tips is the largest engineering community on the Internet
Intelligent Work Forums for Engineering Professionals
-
Congratulations waross on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!
Help with Differential Pressure Reference
- Thread starter ksechler
- Start date
Unless your vent to outside "via a specialty diffuser" is not only impervious to wind effects but also very low resistance, any slight variation in individual room pressures (opening doors, etc) will no doubt alter the header pressure, so it won't be a steady reference after all?
Regards,
Brian
Regards,
Brian
Ksechler
are your intake and exhaust louvres on the same face
If they are, they they will negate any effects due to wind
I would also suggest monitoring the difference across rooms rather than using a general datum.
Generally I have found that inspectors want to know that the difference across a parricular door has been maintained rather than a general reading referred back to some arbitrary datum.
On a recent job we did, we actually did both. i.e. we had differentials across doors and then a general over view (all done in software) It was a pain to commission and the client eventually asked for simple 'across door' values to simplify things.
Why your figures are positive and then negative may be part down to wind, and part to fan control. Also do you have room pressure control dampers fitted. These are essential if you are serious about accurate room pressure control. If they are not fitted (and I have seen some systems without them fited), then the control will be poor and subject to the opening and closing of various doors.
Friar Tuck of Sherwood
are your intake and exhaust louvres on the same face
If they are, they they will negate any effects due to wind
I would also suggest monitoring the difference across rooms rather than using a general datum.
Generally I have found that inspectors want to know that the difference across a parricular door has been maintained rather than a general reading referred back to some arbitrary datum.
On a recent job we did, we actually did both. i.e. we had differentials across doors and then a general over view (all done in software) It was a pain to commission and the client eventually asked for simple 'across door' values to simplify things.
Why your figures are positive and then negative may be part down to wind, and part to fan control. Also do you have room pressure control dampers fitted. These are essential if you are serious about accurate room pressure control. If they are not fitted (and I have seen some systems without them fited), then the control will be poor and subject to the opening and closing of various doors.
Friar Tuck of Sherwood
MedicineEng
Industrial
You don't need to compare with the outside environment. You can simply put your reference in the false ceiling or an exterior corridor. Like that you will have an area protected from wind that is basically stable.
If you want to maintain a differencial pressure system is important to have a door interlock system in place (two consecutive doors should not be able to be opened at the same time).
If you want to maintain a differencial pressure system is important to have a door interlock system in place (two consecutive doors should not be able to be opened at the same time).
Unlike convensional pharma, biotech facilities need to be maintained at lower pressures to atmosphere, particularly decontamination, fermentation and DSP areas etc., to take care of high containment.
As MedicineEng correctly said, it is better to take the atmospheric reference from service areas. I never had problems with this in the past.
As MedicineEng correctly said, it is better to take the atmospheric reference from service areas. I never had problems with this in the past.
It is possible to dampen the outside pressure variations through a device called"Relay Cylinder" which you can buy from control system suppliers such as Sauter.Once you have understood how it works,it wont be too difficult for you to configour something on your own without spending anything!
You have one leg of the DP in the room, and the other in a common header throughout the building right? The pressure in the header is going up and down. I am surprised that different parts of the header are not at different pressure.
Can you use an absolute pressure gauge (reference to vacuum)? You get the absoute pressure in each room, compare this against your barometer (for atmospheric reference pressure), and you will have your DP.
Will this work?
Are you concerned about inter-room DP, or about DP against atmosphere?
- Thread starter
- #8
Ashereng:
Actually we are seeing different pressures in different parts of the header. What causes that? My suspicion is that there is some leakage so air is flowing from higher pressure rooms to lower pressure rooms through the header. Obviously, this doesn't affect the room but it wreaks havoc with the reference pressure.
Do you have any suggestions for an absolute pressure gauge? I need something with a 4-20ma signal. I will use it as in input to our Siemens system and let it do the math.
Actually we are seeing different pressures in different parts of the header. What causes that? My suspicion is that there is some leakage so air is flowing from higher pressure rooms to lower pressure rooms through the header. Obviously, this doesn't affect the room but it wreaks havoc with the reference pressure.
Do you have any suggestions for an absolute pressure gauge? I need something with a 4-20ma signal. I will use it as in input to our Siemens system and let it do the math.
ksechler,
Rosemount 3051 line of pressure transmitters have served my clients well over the years.
They are available in differential, gauge and absolute.
Outputs include 4-20 mA, HART and Fieldbus.
This is my suggesetion, based on my experience. I am not promoting one product/vendor over another, but as a response to your direct request for any suggestion.
Other pressure element and transmitter vendors are too numerous to list, but I have also used products from Yokogawa and Honeywell.
If you google absolute pressure transmitter, I imagine you would get a lot of hits.
As I mentioned above, I would be surprised if you "DIDN'T" see different pressures. Your air hearder is like a pipe. As you go down a pipe, you lose pressure. Why would you expect it to be any different in an air system?
Yes, if you have leaks in your header system, then that would also create pressure differences. But even assuming no leaks, I would still be surprised that the pressure is exactly identical in every part of the header.
If there was no flow, now that would be different. If there is no flow, then I would expect the pressure in the header to be constant in a gas sytem. For the most part, hydrostatic head is insignificant unless there are HUGE height differences - e.g. the top of a 2000 ft mountain versus the base. (In a liquid system, of course you would have hydrostatic head differences at very small height differences, and I digress).
Rosemount 3051 line of pressure transmitters have served my clients well over the years.
They are available in differential, gauge and absolute.
Outputs include 4-20 mA, HART and Fieldbus.
This is my suggesetion, based on my experience. I am not promoting one product/vendor over another, but as a response to your direct request for any suggestion.
Other pressure element and transmitter vendors are too numerous to list, but I have also used products from Yokogawa and Honeywell.
If you google absolute pressure transmitter, I imagine you would get a lot of hits.
ksechler said:
As I mentioned above, I would be surprised if you "DIDN'T" see different pressures. Your air hearder is like a pipe. As you go down a pipe, you lose pressure. Why would you expect it to be any different in an air system?
Yes, if you have leaks in your header system, then that would also create pressure differences. But even assuming no leaks, I would still be surprised that the pressure is exactly identical in every part of the header.
If there was no flow, now that would be different. If there is no flow, then I would expect the pressure in the header to be constant in a gas sytem. For the most part, hydrostatic head is insignificant unless there are HUGE height differences - e.g. the top of a 2000 ft mountain versus the base. (In a liquid system, of course you would have hydrostatic head differences at very small height differences, and I digress).
- Thread starter
- #10
Actually, there really shouldn't be any flow in the header. That's why, theoretically, it should have a constant pressure. Of course, if it's leaking, then there's flow and a delta. At this point I'm getting pretty disgusted with it. I'm going to see if I can instrument it to take the header pressure over time and see how it acts. If it is stable I will add an instrument to monitor the header and use the building control to subtract the offset. If it isn't I'm going to abandon it and just take DP's across doors with the existing instruments.
ksechler,
I wouldn't actually be surprised if there IS flow in your header. Various pipes are connected to it, all of them in turn connected to rooms of varying pressures. There will be a tendency for air to flow from high pressure rooms to low pressure rooms, so the header may indeed have flow through it. If, by chance, it was the "neutral point", then its pressure would remain constant. However, this is probably not the case, especially as room pressures themselves vary, to there is no way the header will be constant.
All things are relative, however, so the variation may be no more than a few N/m2 up or down...it depends on the exact situation.
Regards,
Brian
I wouldn't actually be surprised if there IS flow in your header. Various pipes are connected to it, all of them in turn connected to rooms of varying pressures. There will be a tendency for air to flow from high pressure rooms to low pressure rooms, so the header may indeed have flow through it. If, by chance, it was the "neutral point", then its pressure would remain constant. However, this is probably not the case, especially as room pressures themselves vary, to there is no way the header will be constant.
All things are relative, however, so the variation may be no more than a few N/m2 up or down...it depends on the exact situation.
Regards,
Brian
- Thread starter
- #16
Ashereng and HVACctrl:
Both of your replies were helpful, and are along the same lines I was thinking. The problem isn't so much conceptual. The header is intended to be a common reference for all of the DP's. The problem is the installation. This is a large scale multi-level manufacturing facility, and the header is simply too long and too complex. I think the design should have been for several smaller reference headers to eliminate the long tubing runs.
At the end of the day all I care about is DP between rooms. I am seriously considering abandoning the header system and simply tubing the existing instruments to measure the difference between adjacent rooms. Other options would be to replace the instruments and measure from absolute or completely repipe the header, but I am nearly out of time and money.
Both of your replies were helpful, and are along the same lines I was thinking. The problem isn't so much conceptual. The header is intended to be a common reference for all of the DP's. The problem is the installation. This is a large scale multi-level manufacturing facility, and the header is simply too long and too complex. I think the design should have been for several smaller reference headers to eliminate the long tubing runs.
At the end of the day all I care about is DP between rooms. I am seriously considering abandoning the header system and simply tubing the existing instruments to measure the difference between adjacent rooms. Other options would be to replace the instruments and measure from absolute or completely repipe the header, but I am nearly out of time and money.
MedicineEng, quark, and Brian have comments I agree with (as do others, though). It seems problematic to use a duct manifold as your pressure reference. It's easy enough, as MedicineEng and quark stated, to just put your reference in a nearby service area or corridor.
If you have multiple compartmentalization, DP between the areas is most important, but it should still be referenced to that service area or corridor. This is sufficient for even BSL-4 areas (admittedly with many other factors - linear duct valves on the supply and return ducts will greatly help in this regard).
If you have multiple compartmentalization, DP between the areas is most important, but it should still be referenced to that service area or corridor. This is sufficient for even BSL-4 areas (admittedly with many other factors - linear duct valves on the supply and return ducts will greatly help in this regard).
Absolute gauges seem to be out of question, IMHO, as this is a low pressure requirement and atmospheric pressure may vary by few inches of mercury column depending upon the height of the location above MSL.
The connection of the ambient header to the pressure indicator is dead ended (probably to a diaphragm) and there are less likely chances of leakage into rooms.
Check the resistance offered by the diffuser and also check pressure of the header at various locations with single gauge. There should not be any problem with your existing setup.
The connection of the ambient header to the pressure indicator is dead ended (probably to a diaphragm) and there are less likely chances of leakage into rooms.
Check the resistance offered by the diffuser and also check pressure of the header at various locations with single gauge. There should not be any problem with your existing setup.
quark said:
Could you elaborate a bit on why absolute gauges seem to be out of the question?
I am also a bit confused by the reference to low pressure requirement and atmospheric pressure varying by a few inches. Why would this eliminate an absolute pressure gauge?
I am also not sure what "MSL" is (sorry about that).
quark said:
Please clarify what you mean by single gauge? What are you proposing to check the hearder pressure with?
Thanks.
MSL - Mean Sea Level (reference pressure - 760mm hg)
Absolute gauges are generally referenced to absolute vacuum. The main intention of going for negative pressures in bio contained rooms is to not let anything go into the atmosphere. Suppose, if you have to maintain a 10Pa negative pressure in the area w.r.to atmosphere at sea level, the gauge should read 101315 Pa and at 1km height from sea level, it should indicate 89865 Pa (check my thread on atmospheric pressure vs altitude).
This requires an exact idea of what your ambient pressure is, to do the correct balancing.
Secondly, it is the accuracy of the instrument. Even with a hypothetical 0.001% error, the actual and readout values may differ by few hundreds of Pascals.
So simple way is to go with a DP gauge.
You said in your earlier post that you had observed different pressures along the header. My suggestion is to check the pressures at various places of the header using same pressure gauge.
Absolute gauges are generally referenced to absolute vacuum. The main intention of going for negative pressures in bio contained rooms is to not let anything go into the atmosphere. Suppose, if you have to maintain a 10Pa negative pressure in the area w.r.to atmosphere at sea level, the gauge should read 101315 Pa and at 1km height from sea level, it should indicate 89865 Pa (check my thread on atmospheric pressure vs altitude).
This requires an exact idea of what your ambient pressure is, to do the correct balancing.
Secondly, it is the accuracy of the instrument. Even with a hypothetical 0.001% error, the actual and readout values may differ by few hundreds of Pascals.
So simple way is to go with a DP gauge.
You said in your earlier post that you had observed different pressures along the header. My suggestion is to check the pressures at various places of the header using same pressure gauge.
- Status
Similar threads
- Locked
- Question
- Locked
- Question