Still Puzzled to This Day
Here's a call I went on years ago, and although, conceptually, I get it, it's still hard to comprehend that it really works this way. So, here's the layout:
Computer room at NASA with raised floor and four down discharge fan coil units. Liquid DX chiller with two 30 ton Copeland compressors. A second chiller for 100% redundancy. Computer room did have other equipment added, but was still within the engineered specs for heat load.
Equipment was scheduled to be replaced, but that was going to take a year or so. And one of the top dogs in the A/C service end of things was smart enough to know that if all the same specs were used, they might have the same problem. So I get the call to figure out what is going on.
I walk up to the unit and the water is down to temp. Maybe 36*F, too long ago to remember for sure. Temp split across the barrel was probably only 2 or 3 degrees (maybe even only 1 or 2 degrees), after all, not much of a load. But the room was warmer than it should have been. The fan coils weren't picking up the heat to bring back to the chiller.
I inspected the coils and filters, did all the normal stuff. Then I noticed the circuit setter on the fan coils were all open all the way. That got me thinking. I went back to the chiller and found the triple duty valve. But no access ports. The valve appeared to be open all the way; think I remember something else wrong with it, like no scale on it or frozen.
Told my boss I thought there something wrong with the water flow. Since it was a government install, the original drawings were on file. I got those and went back to the shop to get meter to check flow on the B&G circuit setters at the air handlers.
Found all of the flows to be way over spec'd design. And couldn't throttle it properly at the chiller.
Bottom line, here's what was going on . . . There was too much flow through the fan coil units. Once the water was slowed down, it was able to pick up some heat and bring it back to the chiller. Then the chiller was able to load up and remove the heat, increasing the temp split across the barrel.
As I said in the beginning, it's still hard to comprehend. For example, how does the air going across the coil know how fast the water is moving through the coil? If you have very cold water, regardless of flow, why don't you have cold air at the outlet of the air handler?
Once we balanced the system by slowing the water down, the room came down to temp.
Still baffles me to this day.
By slowing the water down, you increased the contact time the water had with the air passing over it. This is true with dx units too, with water you just need a few different tools.
Remember there is no such thing as "cooling." You are trying to remove heat. The water is the heat absorbing medium. If you rush it through the coil to fast it can't pick up any heat.
Run your finger through a candle flame. Quickly - no problem. Slowly - your finger will pick up more heat and burn.
The original specs will tell you everything.
At this CFM
And this GMP
and this Entering CHW temp
you should have this Leaving CHW temp
Yes, but it is unlikely the discharge air temp changed. Too long ago to remember. So if the discharge air temp did not change, how could the room get cooler with the slower water? Trust me when I say that my boss and I spent a bunch of extra time (at NASA's expense of course) playing with that system and measuring everything we could, and although both of us understand the concepts, it was still quite bizarre.
Originally Posted by ryan1088
The head of facilities was quite pleased and we won the bid to replace the chillers a year or so later.
I completely understand that. What I don't understand is if the coil is very cold, why wasn't the air leaving the A/H cold. It's not like we changed the air speed.
Originally Posted by cjpwalker
Yup, that's why I got the original drawings. Not always possible in most commercial applications when the install is over ten years old.
Originally Posted by pecmsg
Originally Posted by BBeerme
Did you measure the coil temperature? Unlikely that you did or would need to. You probably just measured the air temperature and made an assumption that the coil temp was a similar temperature.
Originally Posted by BBeerme
With water flow, when you exceed the maximum flow rate, you can develop too much turbulence inside the water tubes. This turbulence can create dead spots or pockets where there isn't any any heat transfer. Cold water doesn't mean cold tubes and cold fins...usually it does, but not always.
There may be pockets of super cold air that mixes with pockets of warmer air and you actually measured the average or you happened to get a colder spot. But because the ENTIRE coil probably wasn't cold, the coil appeared small to the load. When you slowed the water flow down and got the ENTIRE coil cold, you made the coil appear larger to the load.
I find that I need to explain this rather frequently. Usually on the other side though. "I thought it would work better with both condenser water pumps running"... not so much. Always refer to design data.
"Trust me when I say that my boss and I spent a bunch of extra time (at THE TAXPAYER'S expense of course)
I corrected the above for you..."
LOL, that's pretty funny! We actually saved the taxpayer money. Maybe you made the assumption that other contractors had never been on that job; maybe you made the assumption that those units were not money pits to begin with. But that's okay. Actually, the units had a long history of service calls due to lack of cooling. The head of facilities needed to find out exactly what was going on.
Since we discovered and corrected the problem in short order, no more money had to be spent on that job for that reason. If memory serves, we ended up replacing at least one if not both of the triple checks. And we had the impeller shaved on both pumps. Which is probably how the problem manifested in the first place.
But it's nice that you have as much concern for the tax payer as we did. We had quite a reputation for solving problems that others couldn't. It's just the way it works is when handing out the service calls, they cannot play favorites. So it's sort of like a back room deal to get us on the problem jobs so we could put an end to the expenses; you know, that trouble call just happens to come to the surface of the pile when we walk in to get another.
Yeah, that job was maybe 20 years ago, so I probably only had about 10 years in the trade at the time. But it was a great learning experience. Here's another one, that proves I learned the value of design data (I'll probably get slammed for mentioning this on a chiller thread, LOL). It was fairly good size air handler for an A/C unit, maybe 15 or twenty tons, can't remember due to the distant time. At some point in time, someone installed hepa's through out the T-bar ceiling. And probably that same contractor installed a bigger blower motor.
Originally Posted by Patrick.hvac
I don't remember the actual motor sizes, but for purposes of this discussion, let's say the original motor was 10hp. And a 15hp was installed with the hepa's. Well, if you look at the fan curve chart for that blower wheel, at that static pressure, if I'm not mistaken, the fan curve didn't just flatten, it actually dipped down a bit. We had to go up to a 25hp motor to get the required airflow and pressure to make the hepa's work proper.
Understanding when you need to look at these data sheets, the job specs, or the original engineering tables can be the difference between gittin' it done right or simply spinning your wheels.
Thant problem worked backwards, according to the good ole btuh=500XgpmXdelta T, increasing flow (gpm) would result in increase heat transfer.
Originally Posted by BBeerme
Bingo. And I'm not afraid or ashamed to say that I'm still puzzled to this day. Every time I think about that job, it still defies logic. The only way I can make sense of it is that we increased the heat of the water back to the chiller, so the chiller was then able to remove it. But did we really increase the total heat back to the chiller? Probably not. You see the circular thinking that develops?
Originally Posted by valdelocc
The only thing that I really know for sure is that the original engineer was right. And the hack and slash mechanics that followed the original install were not right. Maybe the whole problem started when bearings or a seal went bad on a pump. And some contractor said, "if this is good, bigger is better". So, instead of replacing the bearing or seal, he put in a bigger pump.