Modifications to Hydronic System

Santilal

Hi

I posted on this problem about a year ago and have been researching how to fix my system ever since. Basically I have a hydronic system that has in-slab heating, radiators with thermostatic valves and domestic water heating. The system performance is less than optimal and I think my installer lacks understanding. I need to go to him with a solution rather than a problem or I will get nowhere. I can't afford to rip it all out and start again even if that is the best thing to do. So I have four essential problems.



1. There is no hydraulic separation - all circulators interfere with each other.

2. The domestic water heating does not have a circulator and becomes starved of water when the radiator pump is operating. This means I can't have water heating while the radiator pump is going, even if there is spare heat capacity in the system.

3. The two TMV's in the system have a Cv of 1.9. This is too low and needs to be something like 8.

4. The circulators become deadheaded when the thermostatic valves on the radiators all close if there is no other load on.



The attached sketch shows the old system with the proposed modifications pencilled in. Please forgive me for not using the correct symbols. I have not drawn the pressure regulating valves and safety valves, but they are installed. I am asking for some feedback on my proposed modifications - am I on the right track or completely lost.



My proposed solutions are:

1. Installed a closely spaced Tee to separate the boiler loop from the heating load.

2. Install a circulator and balancing valve in the domestic water heating loop.

3. Replace the TMV's with ones with a higher Cv.

4. Use an Alpha2 circulator with bypass valve to regulate the flow to a low value when all thermostatic valves are closed.



Thanks in advance.



Santilal

Gordan

You're not addressing hydraulic separation

Circulators will still compete with each other in your proposed scheme. Why not create a boiler loop (with the thermostatic protection bypass) and pipe each of the (radiator, underfloor, DHW) circuits into the boiler loop with closely spaced tees?

Rich_49

Questions

What type of boiler does this system utilize ?  I cannot determine what in fact is in slab heat from your drawing , others will probably have the same problem .  Are the towel bars in fact the radiators that you speak of and the lower 2 manifolds your radiant in slab ?

  A dedicated circ for hot water is always a good idea , priority for that is also .

  Mixing valves should be as small as possible while not adding too much pressure (head) to system , the less water they have to monitor and mix allows them to be more responsive and efficient . This may be something your installer knows of but does not have a complete understanding of .

Bypasses should be at the radiators so when the TRV is closed your fluid can continue to circulate  and avoid circs pumping against closed valves . 

What make / model IDWH do you have and what are the flow requirements of the other zones ?



  Suggestions :

    No closely spaced tees , use a hydraulic separator , CSTs will further confuse your installer , proper seperators eliminate that understanding issue , they are better also .

Add bypass fittings at your TRV controlled stuff and add a master thermostat set 2 degrees above where your comfort level is in one of those areas with radiators , this will control the circ and it will run as constant circulation .  I prefer Taco Bumble Bees for this as long as the system piping and installed radiators fall within the curve . The Bumble bee also is able to run in constant pressure , set point and Delta T modes . Pressure differential bypass valves are band aids for poorly designed systems !

Size the circulator for your indirect for the capabilities and requirements of that HX .  Install a circulator relay ( 4 zone ) with domestic priority and let it do it's thing .

Taco I series mixing valves in Outdoor reset or setpoint , whichever you prefer , they have a generous Cv so it should be easy to choose . 

Delta T circs on your in floor to optimize performance of each panel

Santilal

Why not create a boiler loop

I am not sure I fully understand. I thought I had a boiler loop separated from the distribution loop with the closely spaced Tee. Could you perhaps provide a sketch.



One of the things not clear on the drawing is that the boiler house is separated from the house by about 15m of pipe. In other words the pipes from the CST's go for 15m before connecting to the rest of the system.



If you are suggesting using a primary loop with its own circulator and using CST's for the distribution system, then I thought that the temperature drop along the primary loop would be too high. Also I was trying to keep the number of circulators down - I thought that too many circs just complicates things. Maybe I just misunderstand your suggestion.



Thanks for the feedback - it all helps.



Kind regards,

Santilal

Santilal

Reply to questions

My boiler is a conventional 30kw (100,000BTU/hr) coal boiler. For a delta T of 15degrees C I need a flow of roughly 30l/min (8gal(US)/min). Hence I need a Cv=8? Is that correct? I agree a smaller valve could be used for the in-slab TMV - I calculate Cv=3 would be okay for a flow of 3gal/min and a temperature drop of 10 degrees C. Sorry for mixing units - we use SI but I know most of you still use imperial.



The in-slab heating is labelled as "U/F Heating" U/F stands for underfloor, which is what it is commonly (and incorrectly) called around here. Sorry for the poor labelling.



The heated towel rails are on the radiator circuit. The radiator manifold was not installed as a manifold. It is in fact a distribution pipe with the individual radiators teed off. The end of the run is just plugged off.



What sort of bypass valve should you use - and am I right to assume you recommend a Bumble Bee for the radiators.



My IDWH (indirect water heater?) is a NZ made one, just made by one of the local manufacturers and supplied by my installer. When I rang the manufacturer and asked questions about ratings and flows etc they thought I was from another planet. But I have managed to identify that it needs a flow of about 2gal/min and this is similar to the in-slab. So I was going to add controls so that when water heating was required the in-slab would switch off. This will not cause any discomfort. The remaining flow of about 5.5gal/min would be required for the radiators when all TRV's are open.



I presume that the separator is fitted at the common point where the piping to the radiators, the water heater and the in-slab come together. I would also then need a circulator for the primary circuit, wouldn't I.



Could you please provide a simple sketch.



Thanks for the advice.

Kind regards,

Santilal

Snowmelt

Nope

The cv ratings are dictated to you as per manufacture specs and are not the same number as g.p.m. I think they are related to pressure drop. From there you have to get all of your pressure drop and size the pump.



It doesn't seem your on the same page as rich in soo many ways. If I was you I get a professional in there to help you.

Santilal

Cv

Thanks for the reply. The reason I am in this predicament is that I have trusted so called experts and I am finding it really hard to find someone who really knows what they are talking about. In order to get my system performing I have spent a lot of time researching books, manufacturer's data and the internet.



On P.398 of John Siegenthaler's book there is a procedure for selecting valves. Step one is to calculate the flow through the valve at design load conditions. Step 2 is to select a valve with a Cv value that is approximately equal to the flow rate calculated in Step 1. This will give a pressure drop of approximately 1PSI at the design flow.



It seemed to me that all valves will cause a pressure drop across them, dependent on the flow. It is up to the designer to determine an acceptable pressure drop. My research has suggested that 1PSI is considered acceptable by the industry. In my particular case, the pressure drop across the valve at design flow is more than the circulator can provide at the same flow. I doubt if my designer actually did any calculations - selection was probably based on what might have worked in the past.



That is how I arrived at my conclusion. Have I interpreted this wrong?



Thanks again.



Kind regards,



Santilal

Rich_49

Cv

the number given by the manufacturer is indeed how many GPM will flowing through while imparting a 1 PSI pressure drop . You are understanding this correctly .

A valve with a rated Cv of 8 will drop pressure in the system by 1 PSI , the other part of this equation is that it will increase the mechanical energy that must be added by the circ (feet of head) by 2.31' .  I will try to free up some time to draw a sketch of what I propose , my schedule is kinda hectic however .

SWEI

Cv and valve sizing

Optimal sizing will deliver full valve authority, which is critical in mixing applications.  We usually aim for a pressure drop of about 4 PSI across the valve at full flow (translation:  GPM = 2 x Cv.)



A few more basic questions: Why are you assuming 15ºF ∆T on the boiler?  What is the design day heat loss for the structure?  Does the boiler have sufficient water volume without adding a buffer tank?

Gordan

No time for a drawing, but

No more circulators than what you have now.



Boiler loop incorporates the boiler, the boiler circulator, the boiler protection thermostatic bypass, and the supply and return "trunks" as you have them now but with two changes: 1) the trunks are connected at the end, creating a loop, and 2) instead of radiator, underfloor and DHW circuits connecting to the boiler loop the way they do right now (supply to the supply trunk and return to the return trunk) they should be connected to the loop each via a set of closely spaced tees, to achieve hydraulic separation. Each of those circuits then has a circulator to induce flow within the circuit. If you put the higher temp circuits upstream of the lower temp circuits the temperature drop won't matter - in fact, it will be beneficial.

RobG

Piping

I think that this is kind of what you are trying to do (see page 43 of attached manual). This is for a completely different boiler than you are using, but it illustrates the piping of a system with mixing valves and closely spaced tees and an indirect. You will have to size you pumps based on your cv's, loop lengths etc....



http://www.lochinvar.com/_linefiles/WH-I-O-Rev%20P.pdf





Rob 

RobG

Coal boiler

Is this a closed or open system? Did you use oxygen barrier pex for the radiant?



Rob

Santilal

Thanks

Thank you for the advice.



I am assuming a 15 Degree C lift in temperature for the conventional coal boiler because the manufacturer advises that the inlet temperature should be a minimum of 60 degrees C and the output temperature is 75 degrees C. Hence a 15 degree lift.



I am the home owner trying to fix the system installed by my installer (who in my opinion does not fully understand hydronic systems and has made some basic errors resulting in poor performance - he is a plumber - the hydronic industry in NZ is not as sophisticated as elsewhere). So I suspect he has not done any heatloss calcs. I certainly have never seen them, but the house is heating ok. It is just some of the flow related issues I am trying to resolve.



The boiler does not have much water but it seems to be enough. There is no buffer tank.



Thanks again.



Santilal

Santilal

Closed system

Hi



The system is a closed system. I am not sure about the piping used but it was a PVC pipe with inner white plastic. It has "PE-Xa 80" written on it so I am pretty sure it is oxygen barrier PEX.



Thanks for the attachment - it is very helpful.



Cheers,



Santilal

SWEI

Kiwi hydronics

now I recall your earlier posts.  15ºC is just fine for the boiler, but I would still do a heat loss calculation before you assume the boiler does not need a buffer tank.

Santilal

Summary

Hi



Thanks for all of the feedback you have given. Based on that I think that I have two options. The first (Option A) is to feed the distribution system with closely spaced tees, and then supply each of the three load circuits each with its own circulator. This means that I will need four circs. The one on the radiator circuit would be a ECM type set to constant pressure mode. See attached drg - Option A.



Option B is to use CST's on the supply loop. Again each load circuit would have it own circ.



The only problem I can see with Option A is that hydraulic separation would not be achieved as all circuits share the common supply piping. This is about 30m (two runs about 15 long) of Rehau Rautitan PE-Xa 32mm his 311 pipe. The pressure drop over this is about 0.8PSI according to the supplier at a flow of 4.5gal/min.



The problem with Option B is that I think I would need to have another circ in the supply loop. I think the load circuit circs would only pump around their own load loop and through the CST, but not around the whole supply loop. This means that there is no mechanism to move the water around the supply loop (but I am not sure about this.) I have not drawn this additonal circ. Also I have rearranged the load to ensure that those loads that required the highest temperature are nearest the heat source.



I would appreciate any comments.



One other question I have relates to the Alpha2 and the Bumble Bee and also the Wilo ECM. If I do use these as proposed they would have to go in the ceiling space. I have read somewhere that these pumps can have a high pitch whine which could drive us all crazy. Can anyone comment on this.



Thanks.



Santilal

Zman

Drawings

You do not need the additional hydrolic separation in your second drawing. The drop in temp between the zones is really not desirable either.



The first drawings is pretty good.

Make sure the main header is sized for 2ft/sec (metric equiv.) max so the circs won't fight each other.

All zone circs should have a check valve.

The bipass valve is not needed or desirable with the alpha circ. It will maintain constant pressure without the wasted energy of bipassing.



Carl

Santilal

Thanks

Hi Carl



Thanks for the reply. From your reply I understand that in my Option B the CST's provide adequate hydraulic separation, but the downside of this is arrangement is that each downstream zone will have a lower temperature than the previous one. Is that correct?



it appears that my Option A is the way to go. Incidentally, this will be easier to implement, given I am trying to fix a poorly performing system.



In Section 8 of the Alpha2 manual it says that you should use a bypass valve, although looking at the performance curves, this seems unnecessary if you use the PP1 or PP2 mode. Would you agree with that? Which mode would you suggest I use - PP, CP or fixed speed?



Also, it seems that the total head available in either PP1 or PP2 is significantly lower than the maximum head available (for a flow of 0.2litres/sec differential head produced is about 2m compared with a maximum head of nearly 6m for the same flow on Curve III for the Alpha 2 L XX60). I think that for my system 2m might not be quite enough. Maybe I need to look at another circ. I think that you have to use an ECM circ for radiators with TRV's to prevent dead-heading. (I have attached the Apha2 Manual).



I will ensure that I draw in check valves on all zones - good advice - thanks.



Kind regards,



Santilal

Zman

Options

You have all the separation you need with option A.

Option B would give you lower temps at each branch. It would also require anther circ.

The bipass shown in the manual is intended for boiler flow protection. You have that covered with the closely spaced tees at the boiler.

All emitters (panel heaters) have a "sweet spot" which is the ideal amount of flow to produce the designed heat output. Once you go past that point, you don't gain very much output for your increase in flow.

I am not familiar with that Alpha and it's performance characteristics as it is not sold in North America.

You absolutely do not want to run the alpha at a fixed speed.I think with your setup you will be best off running it with the relatively flat curve medium head setting of PP2. Once the system is up and running you can measure the supply/return delta t and decide whether you want to change the setting. I think you likely have the right circulator for the application.

Carl

Santilal

Thanks

Hi



Thanks for the advice. I think I now have a plan to go to my installer with. Hopefully he will agree with me, despite the fact that it will point out the flaws in his original installation.



Cheers,



Santilal

Santilal

Installer's Proposal

Hi



Thanks again to all those who have helped me to get to this point. I submitted my proposal based on feedback from this forum to my installer and he has come back with this option. I hate to trouble you, but I would love some feedback. I did have a similar design and I thought it was flawed but I can't remember why. I do think that this option has a nice simplicity about it, but will it work. My comments would be:



But before I start, please note that the electric dump vessel is an additional electric heat to supplement the boiler.



The NPBV on the water heating cylinder does not provide any benefit. When the EMV is closed, the NPBV will bypass water around the cylinder. But the water heating should be treated as a priority load, so why would you ever close this and if you did it would be to presumably make flow available for other loads. If other loads used the flow, then you would not need the bypass valve. I would just size the circ to run water through the water heater all the time - once the water was hot, it would not absorb any more energy.



The lockshield valve on the water heater could be useful for controlling the flow.



There is no boiler protection. Water returning to the boiler could be quite low and cause condensation. The boiler is a 30kW (100,000BTU/hr) coal boiler.



The NPBV on the radiators is a good idea for the times when all TRV's are closed.



Using an ECM circulator could be more efficient and avoid the need for NPBV's.



The circulator would need to be able to deliver sufficient pressure to supply the hydraulic resistance of the system with a flow of 5.7gpm based on a system delta T of 20 degrees C (36 degrees F).



I'm not sure how the in-slab heating will work. I have seen these assemblies before but have never been able to get my head around how they work - I think Watts have one and Caleffi does as well.



There is no hydraulic separation but it might still work as there are only two circs that could interfere with each other. It might be better to use closely spaced tees to supply the in-slab heating.



What do you think?



Thanks.



Regards,

Santilal

Zman

Rory

I do like Rory's plan from a simplicity point of view.

I think your observations are correct.

In reality, the only component of your system that is likely to drag the boiler into a condensing situation is the in floor heat. You could use a smart valve like the taco "I' series that senses the boiler return temp to eliminate that. Closely spaced tees are also a good idea.

The zoning control on the indirect could be done in a way that makes it a "priority", that way the indirect and other zones would never call at the same time.

I think with a few "tweaks" Rory's plan is workable.

Carl

Santilal

I like it too

Thanks.



I like it too because I can put the circ in the boiler house which is remote from the house. I would like to use an ECM circ and I've heard these can be noisy. However, I think this is quite similar to the first install (this will be the third modification!) and the first install did not have enough flow to the radiators. But the boiler loop was not hydraulically separate from the distribution loop in that version.



I am proposing to use a TMV for boiler protection and connecting the distribution loop via CST's. This might improve the flow rate enough to allow the radiators to receive full flow.



I have attached a drawing combining my and Rory's proposal. I am going to meet with the installer next week to see if we can agree.



Thanks for the input.



Regards, Santilal

Zman

p3

P3 is not doing anything in that position.

The radiant loop is being effected by P1

Santilal

P3

The P3 and mixer are part of a proprietary mixer for in-slab. I was going to post separately about it as I can't understand how it works. See attachment.



When you say radiators are "effected" by P1, do you mean P1 is providing the energy for the radiators. That is what the intention was. P1 should drive flow through the radiators, and water heater, and P3 is somehow supposed to draw off just what it needs to supply the in-slab. I am still in two minds as to whether the in-slab loop should be supplied by a Closely Spaced Tee.



I forgot to mention earlier that I will use external control to alternate the supply between the water heater and the underfloor. Basically, the water heater will have first call and when the water is hot the in-slab zone valve will be turned on and the water heater valve turned off.



Thanks again.



Santilal

Zman

Mixer

Thats a cool little mixer.

Yes that will work, It should be piped to the main with closely spaced tees so it is not influenced by P1. I understand that P1 is supposed to push water to the radiators.



Carl

Santilal

Mixer

Hi

How does it work. I see it draws hot water through the right hand connection. Mixed water is drawn from the bottom connection and this mixes with hot water. But won't this mean that the water will now be too hot? Then it mixes with cool return water and this is pulled through the pump. The outlet of the pump then goes three ways - to the in-slab, back to the TMV and the rest back to the boiler. How can you be sure that too much flow does not return to the boiler and there is enough flow into the in-slab. I can't get my head around it - it there a simple explanation?



Cheers,



Santilal



PS I think I now have a solution for my boiler. Thanks for the help.

Tim Potter

concerned about mixer

I was looking at the instructions on the IsoTherm, I have concerns about its appropriateness for your design.



under: "Primary / Secondary piping" "To ensure proper operation and to prevent ghost flow conditions, a primary/secondary piping arrangement is required..."



if I am reading your diagram correctly, thats not how it would be hooked up.



It does appear to be reasonably priced, but an i-valve & Alpha or BumbleBee might be a better choice with ODR & an ECM motor for low cost of operation & steady temps.



Good lock with your re-pipe,



Tim

Santilal

Primary/Secondary

Thanks for that.



I have been thinking about a primary/secondary loop, but then I would have to connect the other two circuits into the loop as well. That would mean I would have to have additional circulators to drive flow in and out of the loop. I was trying to avoid that. I thought that I could have a combined approach but when I drew it up I realised it would not work. I have tried to contact the supplier to see if my connection proposal will work. If not I will have to think some more about it.



The other reason for not using a primary/secondary loop was that there would be a temperature drop around the loop. I don't think my system will tolerate the drop as both the water heater and the Isotherm need 158 degree F water.



It is never straight forward is it.



Cheer,



Santilal