View Full Version : Multi-story fluid distribution

22-04-2007, 10:31 PM
We have a 14-floor hotel that has incrementally reduced the upper stories from the water cooled condenser circuit because of a lack of water flow. They are asking me to bid repiping some of the lower pipe work and to insert balancing valves at each ice machine in the hope of gaining water cooling at the upper floors again. One of my technicians noticed that the upper loop (free loop?) from supply water to returning water was abandoned. And that they have added city water into the upper floors to insure condensing. The piping to and from each ice machine is 1/2-inch copper. The pump seems to be capable of developing 80 psig head.

I am going to try to attach a sketch and I am interested in some opinions.

24-04-2007, 02:33 AM
Also, where is the proper place for balancing valves? In the return or in the supply?

US Iceman
24-04-2007, 04:36 AM
OK Dan, here we go...

The water piping is unbalanced so without any balancing valves the flow takes the path of least resistance, which would be the lower floors.

First of all does the pump have sufficient head capability to reach the top floor plus all of the head losses in the piping?

I would suggest putting the balancing valves on the outlet side of the condensers on the ice machines. If the balancing valves are on the inlet side, the pressure loss of the valves could allow the air to separate and air bind the condensers. If the valves are on the outlet side, then any air released from the water should be taken out by an air separator in the pump suction piping.

Here is some reading material:
http://www.evapcoeurope.net/uploaded-net/documents/eu/WDW%20250-E_1206.pdf (about 2.5 MB)

http://fhaspapp.ittind.com/BG-Literature-Detail.asp?LiteratureID=1195 (about 5.5 MB)

My best guess is the fluid cooler should be in the pump discharge with an air separator in the pump suction. Bell & Gossett used to have some very good design manuals for this sort of thing. You might try contacting them for copies of helpful bulletins also.

it sounds as if this has had problems from day one and the city water supply was added as a "fix" or band-aid to the main problem... insufficient water supply due to poor piping design.

25-04-2007, 02:32 AM
Thanks, Iceman. The supply and return risers are 1-1/2" pipe, with 1/2" takeoffs at each machine. The risers were at one time connected to each other after the uppermost ice machine. I was thinking that reconnecting that link, as well as putting in the balancing valves, would provide better circulation because there would be a negative head available to offset the positive head.

That Bell and Gossett link deserves attention. Interestingly, the hotel just refitted the condensers with Evapco's. But I didn't see any anything helpful there.

I am actually wondering if this damn thing ever worked right. I am bidding the repipe high because I have a feeling that I will inherit some design problems. But I need education on the behavior of centrifugal pumping supply and return pressures. The Bell and Gossett link should occupy me for a while. :)

Ah, the "Ferris Wheel" effect. That is what I am interested in. I should not have to provide a liquid head proportional to the height of the building as long as I have an intact loop. I am thinking that I should reconstruct the continuation of the 1-1/2" piping at the top.

25-04-2007, 08:04 PM
I think your plans are valid Dan.

Looking around the web for some 'inspiration' and found a few items that might prove of use during your thinking/quoting period.

Althoug chemical engineering they have some calculators and design tools that may be useful...

Plus the good old favourite for fluid calcs.

Also something that may be of interest to you is utilising automatic balancing valves as these will aloow for variations in the system if any of the floor have to be shut off for maintenance etc.



25-04-2007, 09:52 PM
14 stories mean 42 to 45 meters.
Pressure increases of 1 bar every 10 meters.
The pressure at the top is 4.2 bar less than at the bottom, and this also means that the minimum pressure to allow circulation is 4.2 bar (please note that US iceman got it: nature spends the least amount of energy, so most of the water will flow in the lower floors).
This circuit is doomed since day one. There should have been 1 circuit every 5 storeys, in order to minimise the pressure difference between the lowest and the highest storey of each circuit.
What you shall bid is 2 complete repiping from ground to storeys 1-5 and from ground to storeys 6-10. The existing circuit shut off at all storeys below 11th, two new set of pumps and etc etc.
You could possibly reduce the circuits to two, from 1 to 7 and from 8 to 14, this means you just need one repiping of the 7th lower floors.

EDIT: are there balancing valves with a kv of over 400kPa?

26-04-2007, 02:34 AM
Brian, thanks for the links. I am intrigued by the autoflow balancing valves. I have much to learn, and you fellows are doing a great job of providing needed info.

14 stories mean 42 to 45 meters.
Pressure increases of 1 bar every 10 meters.
The pressure at the top is 4.2 bar less than at the bottom, and this also means that the minimum pressure to allow circulation is 4.2 bar

Nonick, that's what I was originally worried about until I read the Bell and Gosset pdf that US Iceman sent me. When I referred to the "ferris wheel" it basically came from a chapter that says head does not mean height. As long as you have a complete circuit, the pump head required has nothing to do with how high you are pumping but is only necessary to overcome friction losses. In other words the "height" of the building is taken out of the equation because the return head more or less equals the supply head and cancels the height out of the equation.

I am still sorting this out in my head, but I think I need to reconnect that 1-1/2 piping to properly complete the "ferris wheel". Once I start restricting flow on the lower floors, I feel that a freely flowing loop feed would provide a more reliable supply. I cannot put my finger on it yet, but that's the direction the pros are taking me.

Quoting a repipe of the risers scares me because it is all concealed behind walls and I would have to tear up some significant amount of drywall just to get a feel on how I would go about it. I don't get the feeling that the local engineer is even expecting a price as high as I am already, which is only $30,000.00 for replacing corroded black iron with CPVC.

I really appreciate the thoughtful responses and links. I know a lot more than I did before regardless of the outcome of the bid. And perhaps, I will have an opportunity to discuss it if I am competitive. Thanks again.

26-04-2007, 08:13 AM
head does not mean height

True, if only the system is properly pressurised and the pumps are at low level.

26-04-2007, 06:53 PM
Just a thought Dan, apart from getting the loop going again, do you think that perhaps the system needs chemical cleaning?

Is it possible for scale to have built up inside the system thereby reducing what flow they presently have?

27-04-2007, 03:55 AM
Just a thought Dan, apart from getting the loop going again, do you think that perhaps the system needs chemical cleaning?

I would think that this could be an issue, and I think the engineers at the hotel feel this to be true. If we get the job, we will know when we disable the old piping that is being replaced. The evaporative cooler is an open system and there is a company treating the water with chemicals. I think they worry that a chemical cleaning will deteriorate the walls of the steel pipe which has already begun sprouting leaks in the horizontal runs we are replacing. I am learning a lot. I don't recall seeing any balancing valves between the chiller condensers and the refrigeration condensers. Surely we need balancing valves in the various circuits and not just in the returns of the individual ice machines?

US Iceman
27-04-2007, 04:04 AM
Dan, if the flow is unbalanced you will not be able to adequately determine where the flow is actually going.

Did you say this system was using a closed circuit fluid cooler that was "open"? Or, is this using a cooling tower? An open system and a closed system behave differently...

28-04-2007, 04:32 AM
US Iceman, the water that is doing the cooling is the water that is sprayed on the media which is cooled by the fans. I believe this is a "cooling tower" and not a fluid cooler. Again, definitions and terminology are not something I am confident in when I am looking at things I haven't really studied before.

Do I have my terminology correct, and is this good design, to expose water to open air and recirculate it?

It appeared to me that each chiller had it's own pump and that the refrigeration loop had its own pump. But despite the separate returns, the discharge was a large common pipe. But within the refrigeraiton loop there were several paths on the horizontal level and to a small 2 story degree on another level without any balancing valves visible in the piping.

I feel like I am approaching understanding things sufficiently in order to take a wiser look at things. I probably lost the contract because I priced it rather high for simply running a few hundred feet of Plastic piping, but I felt that I was being asked to do something that some other contractor proposed and I am not comfortable with the proposal.

I have had great input from people. And I have more questions. If pipe fouling is a problem, are the pipes most likely to foul, the ones already receiving the least flow? Can there be a domino effect?

Also, what are the merits of automatic balancing valves. As I study the tables, you size them on predictable flow for a certain pipe size. How is this accomplished? Would I use the same valve for each ice machine as we rise higher and higher?

It's fun to be asking the questions again instead of trying to answer them. :)

30-04-2007, 05:28 AM
The connection you have shown on top is to maintain safe minimum flowrate for the pump and doesn't have anything to do with pipe hudraulics. Like already pointed out earlier, it is difficult to maintain proper flowrates without having balancing valves in the system or reorienting the piping scheme. If you felt balancing valves to be a costlier option, a reverse return header could have provided equal pressure drops across each floor and avoided short cycling.

You do have concentration of salts in evaporative cooling but chemical treatment is also an established technology. Nothing to worry about it much.

30-04-2007, 07:02 AM
This is one I had on my computer but was too large to post here http://www.bellgossett.com/Articles/files/122.pdf
Will have to read thoroughly trough this thread but perhaps a silly idea, if you install refrigerant pressure controlled water valves http://www.ref.or.kr/upload/image_DANFOSS_WVFX_WVS.jpg then each machine will adapt its pressure drop itself till the flow trough each individual machine is enough. If all the valves are wide open to reduce condensing pressure, then this means that your pump is too small.

Perhaps someone said this already but the loop on the top must be closed because the water will take certainly this way (least resistance) unless there's a pressure differential valve mounted in this line.

You need anyhow there or over the pump's discharge and suction line a differential valve in case all the valves on the machiens are closing and the pump keeps running against all valves that are closed.

US Iceman
01-05-2007, 02:32 AM
...the water that is doing the cooling is the water that is sprayed on the media which is cooled by the Fans (http://www.refrigeration-engineer.com/forums/glossary.php?do=viewglossary&term=34). I believe this is a "cooling tower" and not a fluid cooler.

Dan, a fluid cooler has a separate serpentine pipe coil. Water and air flow over this pipe coil providing cooling. The fluid being circulated is "isolated" from the atmosphere. You might hear these referred to as "closed circuit fluid coolers". The name is a tip off to their function. The fluid being cooled is contained within a closed circuit.

A water tower or cooling tower does not have a coil, but does have what is know as "fill". The fill is a simple plastic honeycomb that allows the water to drip down and provides effective distribution and evaporation of the same water being circulated. These devices are "open circuit" as they are exposed to the atmosphere. In this case, the fluid (water) being circulated is also being cooled DIRECTLY by the atmosphere.

There are typically two types of fluid coolers. Air cooled or evaporatively cooled.

An air-cooled fluid cooler is the same as your car radiator. An evaporatively cooled fluid cooler is a water tower with a serpentine pipe coil in lieu of the plastic fill.
Each has it's own set of hydraulic problems with water piping.

Here is some reading material:



This has a good cut-away drawing on the last page of the brochure:

What often appears to be a simple piping problem can become a big mess quickly. If you do not get the project, I don't think I would sweat it. Some of these jobs have a habit of turning nasty as I suspect this one could.