is this configuration viable

timo888

The attached diagram represents my understanding (hopefully not my misunderstanding) of the advice I've been receiving here over the past several weeks, for which I'm grateful.



SITUATION

Old two-storey brick house. No insulation in walls. Old windows. 2" supply and 2" return main lines with 1/2" supply/return branches. A dozen convector-style radiators. All radiators are on exterior walls. ~35 gallons in pipes and 11 gallons in boiler.



Buderus G115/4 (98K btu/hr) with Logalux 53gal DHW tank. Currently set up as two zones with single Taco 007-F5 pump and zone valves.



PROBLEM

By the time the heated water reaches the most remote radiators, the water has pretty much lost most of its heat and those radiators never get warm (they've been bled); the return termperature from the radiator zone is lukewarm at most, maybe 110-115F. The return from the DHW tank is piping hot. I will measure the radiator zone's return temp when I buy a thermometer but by touch alone I'm confident it's nowhere near 130F.



There are a few things that need to be addressed with the system anyway:

-- backflow preventer was never installed on the fresh water feed;

-- feed now goes straight into boiler instead of into supply line below the air separator and the feed valve also seems to be failing (old style Watts iron);

-- no drain/shutoff on the expansion tank



When addressing these issues I want to address the cold remotest radiators and cold return issues too.



I'd leave the old air separator and old feed valve in place, disturbing the installation as little as possible. I would add a second pump and get rid of the zone valves. I would also install (per the advice received here) a 3-way Danfoss thermic valve on a supply-return by pass leg to keep the return temp up.



Is the configuration in the attached diagram viable? Is it missing anything of critical importance? Is anything missing that would make a significant improvement, which would be easy to add and not too expensive ?

Gordan

Some problems

1) You don't need two air eliminators.

2) The makeup supply and the expansion tank should be connected to the system at the same point.

3) This one is a biggie. The way you've got things piped, the thermic valve will not do what you want - it will dead-head the circulators and prevent any flow through the boiler. The boiler circulator has to be between the valve and the boiler.

4) Typically, dirt separators go on the return. The notion is that the system is seen as the source of debris and the boiler is seen as the thing that you want to protect from that debris.



The easiest way to get where you want from status quo is to simply cut in the thermic bypass between the circulator and the (system/indirect) supply tee. Also, tee in the feed line where the expansion tank connects to the air eliminator. The zone valves actually help you.



If you're dead-set on using separate circulators for DHW and space heating, then tee off after each circulator to the supply of the thermic valve, and make sure that you have check valves on both of those bypasses.

timo888

dead-head

Thanks for the critique. Much appreciated.



So the pump must come *before* the tee-off branch to the 3-way thermic. I will relocate them (or relocate it, if I stick with a single-pump system and the zone valves).



The reason for two pumps was that the first one is working but not big enough for the radiator demands in my uninsulated house,so I was going to buy a second pump. Several people have suggested a GrundFos Alpha. If I have two pumps I don't need the zone valves.



But let's say I got an Alpha and swapped it in for the existing pump, and kept the zone valves, so the Alpha was servicing both zones. Would the Alpha's learning mode be able to handle the zone valves and the three possible demand scenarios: radiator zone, the DHW zone, and both zones combined? Or would I have to put the pump into manual mode and pick a speed?

timo888

revised pump location

This diagram shows the pumps ahead of the bypass.

Gordan

Close.

You need two check valves, one on each of the pipes that branch after the circulators toward the thermic valve - the ones you show as "angled". Otherwise, each circulator would induce flow through both circuits.

timo888

I see

Ah. The RADS ZONE pump could get water over to the DHW leg without that second check valve. Thanks.

timo888

possibly stalled return flow from RADS zone?

Most of the details in this diagram can be ignored. The question concerns either a two-pump scenario, or single pump with zone valves scenario, where the return leg from the RADIATORS zone comes down from floors above and meets the return leg of the DHW zone at a Tee.



The temperature of the RADS zone return water is at best lukewarm. (THE BYPASS IS ONLY A PLANNED FEATURE--DOESN"T EXIST YET).



The return temperature of the DHW zone is usually piping hot.



Can the hot return water pumped from the DHW zone stall the weaker gravity-flow of the lukewarm RAD zone, and prevent the proper circulation of the RAD zone, when the two zones meet at the return Tee circled in red?

Gordan

In a word, no

You're using circulators, which means that you're not relying on gravity flow.

Ironman

Suggestion

Your diagram looks OK if the circs have internal checks. If not, you'll need two flow checks.



My suggestion  is this: Since you've had so many issues with your system, why not post a diagram and photo's of what is actually there and let the pro's here review it and make suggestions as to the best way to correct it. Even though your diagram looks right, it may not be the best way to correct the problems. For instance, if you install a bypass or mixing valve on the boiler, it will take longer to get heat to the rads. You've already mentioned the problem of getting heat to the farthest ones. A bypass will probably make that worse. Do you have valves on the rads? If so, they could be used for balancing. If not, you may need to look at piping the system in reverse return or primary/secondary with the the zone loop reverse return.

timo888

gravity

I don't have a really firm grasp of pump concepts yet. If there's a 1" diameter pipe going straight up 100' towards the sky, how much head does the pump need to get water to come out of the top of the pipe? If the correct answer is anything less than 100' then I am in the dark, which may very well be the case.



My question above is meant to explain why I thought the rad-zone return flow could be "weaker" than the DHW return flow. I am thinking that the pump head gets "exhausted" or diminished when it has to push the water up a couple storeys vertically. Or does the pump constantly rev up in order to maintain a constant pressure throughout the system?



Also, somewhere along the way I got the impression that hotter water flows faster than cooler water, and the DHW return is very hot whereas the rad zone return is relatively cool.



Does the pump handle all of these variables and maintain a constant pressure in any case, so that if the DHW zone is having water pumped through it and the rads zone is having water pumped through it BY THE SAME PUMP (to keep it simple), when the two return flows meet at the boiler return Tee, the two flows will just merge, and one is not stronger than the other and "hogging the pipe" so to speak?

timo888

sketch

That would be great.



It's hard to show it in pictures since all the boiler piping is between the boiler and the wall and there's no way to get a panorama view, and the side view shows very little. But I can sketch it.



The rads have no adjustment valves.



I don't know what this is: "primary/secondary with the the zone loop reverse return."



I will take a crack with this: http://sketchup.google.com/



P.S. I could always wear a sweater. My main goal is to make sure the boiler isn't being damaged or having its useful life shortened by return temps that are too low.

timo888

direct return rad zone(s) sketch

Couldn't figure out how to do turns with pipe on the google sketching software, so here's a crude sketch.



The mains are 1-5/8" O.D. pipes. 1/2" pipe for supply/return to each rad.



At present there are only two zones coming out of the boiler (DHW and "All Radiators") serviced by a single pump with two zone valves, but the house was plumbed with a rear-of-house rads zone and a front-of-house rads zone.



The length of each 1-5/8 main loop is 96' (12' + 24' + 12' out + 12' + 24' + 12' back) . So almost 200' total for both loops combined, not counting the 1/2" supply/return to the individual radiators. It's probably 4' of 1/2 pipe for each first-floor radiator (24' total for the six rads) and 20' for each second-floor radiator (140' total for the 7 second-floor rads). So say 164' of 1/2 pipe.



About 13 gallons in the main loop (assuming 1-1/4" ID). Aboout 6 gallons in the 1/2 lines (assuming 1/2 ID?). Just shy of 11 gallons in the boiler. So about 30 gallons total. We have a Taco 007-F5.



Radiators #1.4 (first floor powder room) and #2.3 (second floor master bathroom) and #2.4a (second floor rear bedroom) stay cold. When bleeding these radiators only a small dribble of coolish water comes out, whereas the radiators closer in to the boiler deliver a stream of hot water, the closer in, the more forceful the stream. Would that mean the Taco 007-F5 pump is undersized for the direct return plumbing?



http://cbs.grundfos.com/CBS_Master/lexica/HEA_2-pipe_system.html#



[quote]Two-pipe system with direct return

In a two-pipe direct return system, the total pipe length from the pump to and from each radiator is shorter for the radiators closer to the pump and longer for the more distant radiators. For this reason the differential pressure can be significantly higher at the closest radiator than at the most distant radiator. This must be taken into consideration when designing the system.[/quote]



How do you "take it into consideration"? What should be done differently?





I accept the point that a bypass could cause even less water to be sent out to the remotest radiators. Would replacing the pump with a more powerful one, so more gallons per minute are moved through the system, compensate for a bypass?



If a more powerful pump would help, I'd like to get one with the same relatively flat curve as the 007, because of the zone valves.

Ironman

Sketch

Tim,



Are the supply and return lines connected at the ends as your diagram is showing?

James Day_2

rads

I'm wondering have the radiator valves been replaced?  Are there TRV's on the system.  Sounds like a balancing problem.  Could have stuck or clogged radiator valves not allowing flow to those rads.

timo888

not connected

Bob,

I've edited the diagram. Sorry about that. And here's a picture of how the supply from the boiler transitions from copper to iron pipe, showing the front and rear legs, and the returns in the background.

James Day_2

returns

Is it my eyes or are the returns small.  Is that 3/4 pipe?

timo888

probably original

It's remotely possible that the convectors could have been replaced at some point, but they look original to me. I know they haven't been replaced in the last 30 years. We've been here 20 years and the former owners were here for 10, and they did nothing but some cosmetic touchups here and there.



There are no valves on the individual RADs. They're connected to the 1/2" supply and return pipes with unions, and there's a pipe with bleed valve that rises about 18" from the convector.

timo888

pipe diameter

The returns are 1-1/4 (1-5/8 O.D.) but transition to a smaller diameter pipe right before they come together and transition to a single 1" copper pipe.

Charlie from wmass

Is this a converted 2 pipe steam system?

was this steam and converted to hotwater?

Ironman

Thanks Tim

Could you also post a pic or two of one of the problem rads and their connections to the mains? Make sure the piping at the rad is showing, please.

Ironman

Charlie...

I was having the same thought.

CMadatMe

Why The By-Pass

Get rid of it. There is no need for it. Boiler can handle 115-120 degree return water temp. I would be more concerned about raining in the chimmney then protecting the boiler. Are we sure that flat headed curve pump can do the job? Your getting too caught up in trying to make a simple piping job into one you don't need. This is a system that should be running on constant circulation. Would use a Taco VDT over the Alpha. Set the delta-t on the pump, set the ODR curve you want and let it fly.

timo888

piping to rads

We need to remove the metal cover and stick the vacuum hose under these convectors a little more often :-p

timo888

but what if return temp is not reaching 115?

My kids' foreheads when they've been feverish are hotter than the return pipe is in midwinter. If I put the heat on now the return pipes would be piping hot. But not when it's 10 degrees outside. Then, they're lukewarm. I'd be surprised if they were reaching 115. There is no insulation at all behind the brick, and the windows above the radiators are drafty.



No, 'm NOT sure the 007 can handle it. Thanks for the recommendation of the VDT. I will read up. I like TACO as a company. Excellent website resources. And when I''ve phoned they've been very helpful with information. Thumbs up to Taco.

timo888

steam ... maybe

I was told the old boiler was originally a coal burner that had been converted to oil. House was built around 1946 if that's any indication of whether it could have been a steam system converted later to hot water.

Charlie from wmass

If it is converted from steam

I wonder if there are orifices still in place?

Ironman

Delta T Too High - Gpm Too Low

After seeing more pics and info posted, I believe your problem is that your not moving enough gpm. The fact that your returns get hot now but not when it's cold would definitely point to this.



I agree with Chris that a variable speed delta T circ might be the answer to your problem. If this is a converted steam system, the 1/2 in lines going to your convectors may be restricted and that would cause the high Delta T, low gpm condition that you have described.



I have never had to install a bypass or mixing valve on a Buderus - they just don't need it. That's right from my rep's lips. And I thought that you had posted that you had lined the chimney. If not, that should be done if needed.



If you go with v.s. circ, you'll need two circs; one for the indirect(the 007), and the v.s. for the heating loop. You could also split the two heating loops into two zones by putting a zone valve and t'sat on each.



My $.02 worth. Maybe someone else sees something else.

CMadatMe

Nothing Else To See

Except to take a step back and get back to basic hydronics. What's the heat loss, what gpm do we have to move and then what head do we have to overcome. Then on to a simple standard piping arrangement. Boiler is set to handle the application. He just needs to keep it basic. System doesn't require anything more then that.



GPM is easy. Simple hydronics formula. For a standard 20 degree delta-t, take the heat loss divided by 10,000 your on your way. For pump head start with the simple rule of thumb formula for estimating head.



Longest Run From the Outlet side of the circ thu the farthest rad back through the boiler to the inlet side. Multiply times 1.5 then multiple by .04.





100 x 1.5 = 150 x .04 = 6 or 6' of head...



Its a starting point.

timo888

advantages of variable speed?

model 7: http://www.taco-hvac.com/uploads/FileLibrary/101-029.pdf

VDT: http://www.taco-hvac.com/uploads/FileLibrary/100-68.pdf

3-speed: http://www.taco-hvac.com/uploads/FileLibrary/101-125.pdf



The simplest least expensive upgrade would be to swap in a pump with the same length dimension as the 007 [6-3/8 or 162mm]. No new plumbing and no additional wiring would be required.



The VDT would require replumbing because we'd have to dedicate a pump to the rads zone and another one to the indirect.



The 3-speed pumps are size-compatible with the 007. The 0010 seems to deliver greater flow than the 007 when it's in "high" speed mode. But when reading about the variable speed I found this:



[quote]There's no risk of over-sizing a 00-VDT circulator, even if actual installation varies from the original design. By maintaining a constant Delta-T, the 00-VDT circulator automatically varies the flow as needed to ensure optimal performance and heat transfer.[/quote]

http://commercial.taco-hvac.com/news.html?action=detail&newsid=492



Can oversizing the pump damage the system? If so, what's the margin of safety here?



If the lone 3-speed pump in high mode did not diminish the midwinter delta T and increase the temps of the farthest rads, then we could always move to the more sophisticated upgrade, replumbing for two separate rads zones and getting a third pump.

CMadatMe

Pump Curve

How can you determine pump size without knowing gpm requirement and head? What is the total system required gpm? What is the head your trying to overcome?

timo888

pump head

What I've simply been looking for is a pump that delivers more GPM than the 007 we have now (and with roughly the same gradually declining curve "shape" as the 007). But maybe it isn't simple and the GPM the pump will deliver is dependent upon the maximum pressure the pump attains given the resistance of the plumbing?



With my limited understanding the GPM requirement is just this: "greater flow than we are getting now with the 007". I cannot put a number on it. I am still trying to understand your earlier post -- I am ignorant about pumps, and calculating pump head is a skill I do not have. I will have to do a lot more reading.



Is calculating head for a direct return two-pipe system different than calculating it for other plumbing schemes?



Knowing that I don't know the answers, the 3-speed seemed a good way to go because it offered some ablilty to tweak the performance manually, and if I had to resort to the more sophisticated "Plan B" it would not have to be tossed but could be incorporated into the solution.



If I'm reading the chart correctly, at 6' of head the 007 delivers 3 m^3/h of flow and the 0010 in high-speed-mode delivers 4 m^3/h of flow. The 0010 chart doesn't express it in GPM, but 4 m^3/h is about 18 GPM and 3 m^3/h is shy of 14 GPM.

CMadatMe

Heat Loss Heat Loss Heat Loss

Do a heat loss and learn this. You will then know how to calculate gpm. If you cannot define greater flow you cannot pick the correct pump.

http://flopro.taco-hvac.com/video.html?id=44



Take a measurement of the longest run. From the outlet side of the circulator thru the furthest emitter back down the return, thru the boiler to the inlet side of the circ. Let's say that total run is 150'.

150 x 1.5 (for fittings, vlvs, etc) x .04 (ft per sec.) = 9' of head.

This is a simple rule of thumb formula for estimating head.

timo888

BTUH

Per the Taco video, BTUH= "the heating requirement".



If this number is different from the BTUH rating of the burner (e.g. 98000) how is it determined? Do you start with the volume of water in the system (pipes + boiler itself) and calculate the number of BTUs required to raise that amount of water to 180F?

Ironman

Idealistic vs Realistic

Idealistic: If we were designing this system or replacing the boiler, then a heat loss would be mandatory along with calculating piping loss, gpm, etc.



Realistically, we're trying to correct a deficiency that may be due to incorrect pump sizing, restricted piping, or both. If we had new piping we could easily calculate the head loss. But we don't know the condition of these old pipes and I highly suspect the 1/2" run outs are restricted to some degree. Therefore, a piping calculation is only a guess at best. Better than nothing, but still guessing.



We know from the boiler size that we should be moving 8 - 10 gpm. And if there is any restriction in the piping, then assuming 10' or more of head is not out of line. The 007 is obviously not doing the job.



I would suggest trying the 010 3 speed. It's the simplest thing to do to see if it would correct the problem.



I would also at the same time remove one of the problem convector elements and check for orifice plates or crud restricting it. You could take it outside and connect a hose to it and see if you get good flow through it. Also, look inside the 1/2" pipes feeding it and see if they're restricted. Once you've established that there are no restrictions, then install the 3 speed pump and begin on low speed to see if your problem is resolved.

timo888

orifice plates

Will an orifice plate always be installed between two flanges and have a little handle? Our cold rads look to be connected via standard pipe unions.

Ironman

Orifice Plates...

Would be in the unions of the rads. You may not have these, just a possibility.



I'd also check your circ - it may have debris in the impeller.

Mark Eatherton

Orifi...

and gravity systems.



http://www.oldhouseweb.com/how-to-advice/gravity-hot-water-heating-continued.shtml



Scroll down for a picture of a restrictive orifice in a radiator union.



ME