Page 1 of 3 123 LastLast
Results 1 to 13 of 31
  1. #1
    Join Date
    Jan 2003
    Location
    USA
    Posts
    5,197
    Post Likes

    What’s the tightest temp spec you have controlled?

    Was talking with a customer’s ME today on an upcoming renovation that’s in the early design phase. They mentioned they sat down with the group that manages the existing space to figure out their requirements so they could make sure whoever they hire for the design is clear on what’s required. They want +/-1F, which I said was going to be a costly headache. Talking about it more it comes out they want no more than 0.3F temp change in an hour. This is their key requirement. Back of mind I’m thinking, is that even possible? Add a few occupants in the space, fire up a bit of equipment whoops your out of spec. Hot fart in the room, out of spec again. Air not perfectly distributed in the space... Partly cloudy day and the sun hits the building between clouds, there another nasty email chain of complaints. Haul in whatever materials are being worked on in the space...yep lost it again. Asked if they had done any research on other facilities that may have pulled this off, not really. Do the managers have their facts/requirements correct? Yep some crazy machine has this spelled out by the supplier. Ok then... Get your checkbook out, that’s going to require high end equipment to have a chance of getting there.

    Glad I’m not the poor bastard that has to come up with the mech design.

    So short of NASA environmental chambers has anyone seen or heard of this kind control pulled off in a commercialish type space? Smithsonian, crazy process area, etc maybe? This is not a sealed chamber, this is an area people will be coming and going, has some large equipment, a “general” work area in a commercial facility.

    Have a done a few projects with some tight requirements, but few tenths of degree is 10x more than I have heard of. If they have their crap straight, I can't wait to see what kind of HVAC they toss at it.
    Propagating the formula. http://www.noagendashow.com/

  2. #2
    Join Date
    May 2014
    Location
    Bay Area California
    Posts
    26,599
    Post Likes
    Calibration labs do it frequently. Not really that big of a deal. Cooling runs 24/7 and the heat strips are controlled to maintain the constant temp. Self adjusting controller to change the PID's as required.
    I do a triple evac with nitro to remove non condensables.

  3. #3
    Join Date
    Jan 2003
    Location
    USA
    Posts
    5,197
    Post Likes
    Thread Starter
    What are they using for the space sensors?

    Even the high end Vaisala gear I would normally has an accuracy of +/-0.18F not sure what the repeatability is. Could be a mile off as long as it can repeatedly get exactly back on the same spot every time. There is also an additional 0.1%FS error on the analog output even after calibration. Normally want my key sensors 10x better than what I'm trying to maintain.

    Quote Originally Posted by BBeerme View Post
    Calibration labs do it frequently
    That's the comical thing. The onsite cal lab requirements are a joke to what these guys are asking for. Guessing they must not get their wiz bang bits calibrated internally.
    Last edited by orion242; 03-12-2020 at 07:38 PM.
    Propagating the formula. http://www.noagendashow.com/

  4. #4
    Join Date
    May 2014
    Location
    Bay Area California
    Posts
    26,599
    Post Likes
    Not so sure the sensors are the critical part, more like the controller that is able to change the rate of the pulsing heat strips. And to adapt the PID values to know when to pulse more or less depending on the rate of change. And which direction. Obviously, an integral part of this whole system is good air flow to mix the air properly.

    The last cal lab I was in had an incredibly tight temp control. Everything had to stabilize at room temp for 24 hours before any measurements could be made. They measure size. So the expansion and contraction due to temp fluctuations are measurable.

    They had their own sensors in the room to track the temp. Room was maybe 40' x [80 to] 100', just a guess, been awhile since my last call there. Anyway, they were tracking the temp then an alarm triggered, which prompted my last call there. They had two of their own sensors close to the two sensors (averaged) that controlled the pulsing heat strips.

    I went in and told them everything was working properly, at least from what I could see. Now here's the funny part:

    They showed me their charts and graphs. Insisted, since they were a cal lab, that their sensors were perfectly calibrated. I asked the two guys in charge to follow me into the cal lab. Then I pulled my un-calibrated pocket thermometer out of my shirt pocket and measured the temp at their two sensors. Which near perfectly the same. Then asked them, how could that be? Your sensors say there is a 'great' temp difference.

    Come to find out, one of their sensors went bad. The head guy was so embarrassed, he never spoke to me again, LOL.

    Don't have a clue what type of controller they are using. I knew once, but that was maybe two years ago.


    Quote Originally Posted by orion242 View Post
    What are they using for the space sensors?

    Even the high end Vaisala gear I would normally has an accuracy of +/-0.18F not sure what the repeatability is. Could be a mile off as long as it can repeatedly get exactly back on the same spot every time. There is also an additional 0.1%FS error on the analog output even after calibration. Normally want my key sensors 10x better than what I'm trying to maintain.



    That's the comical thing. The onsite cal lab requirements are a joke to what these guys are asking for. Guessing they must not get their wiz bang bits calibrated internally.
    I do a triple evac with nitro to remove non condensables.

  5. #5
    Join Date
    Jan 2003
    Location
    USA
    Posts
    5,197
    Post Likes
    Thread Starter
    Quote Originally Posted by BBeerme View Post
    Not so sure the sensors are the critical part, more like the controller that is able to change the rate of the pulsing heat strips. And to adapt the PID values to know when to pulse more or less depending on the rate of change.
    These are non-issues, do this on other spaces with demanding requirements. More on the +/-3F 3%rh type stuff. However if I can’t read at least maybe half of spec I need to maintain, recipe for failure IMO. Need to see which direction to go before well before running into the wall. If three tenths is the span of the requirements, ideally I want my sensor to read hundredths so I can react before I get more than a tenth or so from the upper/lower bounds.

    Doubt they would use any PWM type control but who knows. Year-round steam, HW/CW are available.

    Quote Originally Posted by BBeerme View Post
    Obviously, an integral part of this whole system is good air flow to mix the air properly.
    No doubt about that. Will have multiple sensors in the space just to CYA against any gradients in the area if and when it gets to that point.

    Quote Originally Posted by BBeerme View Post
    The last cal lab I was in had an incredibly tight temp control. Everything had to stabilize at room temp for 24 hours before any measurements could be made. They measure size. So the expansion and contraction due to temp fluctuations are measurable.
    That's what's going on here. From the gear I saw, extremely accurate dimensional measurements. Least what I could identify. Given the size of gear and what they could fit on it, large objects are going to be moving through the space. So, does one have an acclamation room adjacent to the measurement space? Not rolling in some 1k# object and expecting <0.3F/hr change till that object is at the same temp as the measurement area. Even occupants entering the space in the morning and firing up anything that’s not already running 24/7.

    Right now the HVAC in the area is 70s era dual duct unit with pneumatic bunson boxes and newer single duct vavs scabbed on it. 0.3F/hr lol...never happened in the existing space and boy have I heard about this till I explained the reality of the situation.

    Quote Originally Posted by BBeerme View Post
    They had their own sensors in the room to track the temp. Room was maybe 40' x [80 to] 100', just a guess, been awhile since my last call there. Anyway, they were tracking the temp then an alarm triggered, which prompted my last call there. They had two of their own sensors close to the two sensors (averaged) that controlled the pulsing heat strips. .
    Pretty much what I deal with. They have some networked Fluke temp/rh device in any space they care about the conditions for monitoring. Anything the BMS does is compared to those. Seen many, many a log from those and the Vaisala’s we use seem to put them to shame. Looked at the spec for these Fluke jobies and its was on par with Vaisala’s spec. All these Fluke devices are on a cal schedule and that is handled internally. They are bang on every time I get logs jammed in my face, they just seem to be taking an average over time or some other filtering when I lay their data over ours.

    If they are serious about this requirement, thier Fluke jobbie and what I typically use isn't cutting it IMO.
    Last edited by orion242; 03-12-2020 at 09:14 PM.
    Propagating the formula. http://www.noagendashow.com/

  6. #6
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,905
    Post Likes
    I will be disappointingly vague on some details, Orion, my apologies.
    I had a site with a +-.5 deg space. They used a dedicated AHU that had a fan that turned the entire space into a planum. One wall was perforated as a supply, the other entire wall perforated as a return. The air in the space changed at something like once ever 10to 30 seconds in a low velocity, high volume way. The room was essentially a fat duct. You could slow but purposefully walk the room and that was how long the air was in there.
    CHW to get the cooling/humidity. HW to get the space temp right on and allow for dehum.

    I do not know what they were using for space temps or even what I was using for control, it was 10 years ago almost. I do remember we had never used the sensors before, and we had a calibration agent calibrate our inputs electrically using a decade box. He was good, very helpful. My apologies. I will say relying on a dual duct AHU from the 70s is a non starter. He needs his own system.

    The space had a hiccup when the mech engineer placed the supply fan motor in the airstream and the cooling system could not handle the heat load of the motor, they had to renovate to get the motor out of the airstream of the room... which was fun for them....

    I remember the cooling load to be in the 3-5 ton range, not huge. Just GOBS of airflow. The hot farts were evacuated IMMEDIATELY. The space temp.was the DAT. Easy for me to control, hard for the mech guys to impliment.
    I hope that helps!
    Hmmmm....smells like numbatwo to me.

  7. #7
    Join Date
    Jan 2003
    Location
    USA
    Posts
    5,197
    Post Likes
    Thread Starter
    Quote Originally Posted by numbawunfela View Post
    They used a dedicated AHU that had a fan that turned the entire space into a planum. One wall was perforated as a supply, the other entire wall perforated as a return. The air in the space changed at something like once ever 10to 30 seconds in a low velocity, high volume way. The room was essentially a fat duct. You could slow but purposefully walk the room and that was how long the air was in there.
    Kinda what I was thinking. Can't see how else you could pull this off.

    Quote Originally Posted by numbawunfela View Post
    I will say relying on a dual duct AHU from the 70s is a non starter. He needs his own system.
    All the existing equipment is clapped out garbage that's been completely sodomized with 30+ yrs of bandaids. This will be a new space, all new equipment. Would have left laughing if they proposed anything less.
    Propagating the formula. http://www.noagendashow.com/

  8. #8
    Join Date
    Jan 2003
    Location
    USA
    Posts
    5,197
    Post Likes
    Thread Starter
    Thanks guys.

    Guess what I’m hearing is it is possible which was the first concern. Should be a fun hunt for sensors when the time comes.
    Propagating the formula. http://www.noagendashow.com/

  9. #9
    Join Date
    Jan 2010
    Location
    Sea to Sky
    Posts
    3,673
    Post Likes
    Just wait until they start telling you they want champagne but only have a cooking wine budget......

    Sent from my SM-G965W using Tapatalk

  10. #10
    Join Date
    Jan 2003
    Location
    USA
    Posts
    5,197
    Post Likes
    Thread Starter
    Lol.

    Should be an economic theory. As the decimal point moves to the left on your requirements, it moves to the right on my quote.
    Propagating the formula. http://www.noagendashow.com/

  11. #11
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,905
    Post Likes
    The spec is so tight that the things you will be watching out for would seem to be
    The valve authority on the coils you got. No tolerance for even slight hunting. You may need to do a 1/3 - 2/3 setup on each coil to expand the amount of signal resolution you have on each coil. Common on steam, but not unheard of on CHW HW applications. I know you are exploring signal resolution on actuators now, so I am sure you are picking up what I am laying down...
    Or using a 3 way valve on the coils allows you to modulate the temp of the coil precisely. Then give the coil a dedicated circulator after the 3way to make the flow super high. Then the temp in the coil is consistent accross the entire coil because of the high flow of the dedicated actuator. No entering water at 120, and leaving water at 80, and an inconsistent coil temp accross the face.
    Also my room had perforated walls on the small walls, the long ones had windows and doors. So air travel in the room was longer. Your room may need to keep the amount of time the air is in the room be either keeping the room small, or perforations the long walls, making the amount of travel shorter.
    Also placing the processes against the supply side wall to get the tightest control on the processes, and the slopiest control at the side of the room with no processes on them. Then if control is +-.08 on the supply side and +-.14 on the return side of the room you are still good.
    I tend to hope for the best and plan for the worst on these.
    There would need to be baffling in the supply plenum to ensure consistent output accross the face of perforated wall. Not all the flow on the left or right or whatever side the supply duct enters the plenum behind the perforated wall. The fan will likely see high static between sup and ret, but the room will feel just normal.
    Also monitoring static between the room and the rest of the space to keep the room SLIGHTLY positive so that you do not suck air into the room when the door opens - and a dedicated anteroom of course.
    I do not think I need to tell you that stressing early and often that you can control anything, but you can only meet spec with good mechanicals helps set expectations.
    Just a few ideas as I spit ball the design in broad strokes. Hope it helps. You do not do the mechanicals, but you can make suggestions that, if not taken, you can point to later if it all goes south at the very least.
    Hmmmm....smells like numbatwo to me.

  12. #12
    Join Date
    May 2014
    Location
    Bay Area California
    Posts
    26,599
    Post Likes
    Here's a couple of simple things I remember about these rooms, the first is real simple. Lights stay on 24/7.

    This next one can be a real stickler, and it may be what Numbawun is speaking to above. I think the term is Gradient. The gradient in the room is the temp difference from one place to another. Be real careful if there is anything like that in the specs.

    The first time I had experience with this was a large environmental chamber. Turned out only about 50% of the box met spec, pretty much right in the center. Another time, in a small room, the temp was all over the map. So they decided to drop that spec, probably would have cost them tens of thousands to get a consistent gradient throughout the entire space.
    I do a triple evac with nitro to remove non condensables.

  13. #13
    Join Date
    Sep 2007
    Location
    Kenilworth NJ
    Posts
    1,905
    Post Likes
    Quote Originally Posted by BBeerme View Post
    I think the term is Gradient. The gradient in the room is the temp difference from one place to another. Be real careful if there is anything like that in the specs.
    What he said... hehe
    Hmmmm....smells like numbatwo to me.

Page 1 of 3 123 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  
Contracting Business
HPAC Engineering
EC&M
CONTRACTOR