When the Controls Just Won’t Work

I have an old engineer friend I’ll call Bucky (not his real name). Bucky was burned years ago when he designed an HVAC system that turned out to have insufficient capacity to keep the building comfortable in winter. In fact, the perimeter offices were in the low 50s (°F) (low teens °C) when the first cold snap hit. When I say he was burned, I mean it figuratively – but the occupants of the building were thinking about burning him literally.

Well, old Bucky was not going to be burned again. We joke that the architect has to specify stronger door hardware when Bucky is doing the mechanical design, so that the doors don’t blow off the hinges from his absurd supply air quantities.

This leads to a control problem on Bucky-designed jobs. I had to install and program a building automation system for a Bucky job, and it wasn’t a good experience. How can one tune an office temperature control loop when the reheat box can warm the room faster than the temperature sensor can respond? The occupants would essentially be subjected to supply air temperature, which could reach 130°F (54°C) in heating mode.

I first went with my tried-and-true PID tuning method that I learned from a DuPont instrument engineer in the early 1980s. This method had never failed me until I tried it on Bucky’s HVAC system. I worked for a couple of hours on a single office but I could not get anything near stable control. I tried adding a feedforward loop to give the PID loop advance notice that the oversized hot water valve was about to open. That took programming time and it didn’t help at all.

So I went back to the office and batted the problem around with a group of my peers. We discussed, we calculated, we got out our controls books. We came to no good conclusion.

The next day, I programmed all of the interior spaces with no problem. There was way too much air but my tuning method resulted in stable control on the first pass. Then I went to ponder the perimeter offices again. As the building was approaching occupancy time, the painters were gone and the carpets were being installed. There happened to be a carpet layer in the office I went to first. I thought out loud for a minute, then I vented to him about what a pain the air system was for me. He sat up on his heels and listened, then said, “Seems to me there ought to be a way to reduce the air and water flows.”

I turned mighty red with embarrassment at that time. I thanked the carpet layer and went to call the test and balance fellows. They agreed to cut back on the water and air to the perimeter spaces if I could convince Bucky that I needed it. Well, Bucky came to the job site and it was pretty easy to convince him by getting him to stand in a perimeter office for a while. Problem solved, but not by me. They say that if your only tool is a hammer, then every problem looks like a nail. That was my problem in a nutshell. I was a controls guy, so I focused only on the controls.

The lessons I learned were: 1) Engineers can be wrong (yes, really!); 2) When a system can’t be tuned, the system might need fixing; and, 3) When all else fails, ask the carpet guy.

1 thought on “When the Controls Just Won’t Work

Leave a Reply

Your email address will not be published. Required fields are marked *