gravity conversion & flow rates

R. Kalia

I read the article on gravity conversions and flow rates:

http://www.heatinghelp.com/newsletter.cfm?Id=125

But if I want ΔT=20, the flow rates listed for gravity conversions are way too high. I calculated this using the specific heat of water, see below and please check if possible.

My house was designed for around 160,000 BTU, so the chart says 25 gpm for a gravity conversion. Water specific heat is 8.34 BTU/gal/F. So heat delivered = 25*8.34 BTU/minute/F = 12,400BTU/hour/F. so ΔT = 160,000/12,400 =13F.

Further, for the actual BTU needed on a design day (only 80,000 because of insulation), I would have ΔT=6.5, which is wasteful and would reduce efficiency in a condensing boiler.

The article says people have flow rates that are too high, but I claim that even the flow rates listed in the article are too high. The flow rates listed for 'modern' systems give closer to 20F. But if the actual BTU is much lower than the design BTU, then even in modern systems the ΔT is much smaller than 20F.

How should I determine the flow rate? ΔT=20, or the article?

Alan(CaliforniaRadiant)Forbes

A very good question

You should determine flow rate based on the article for the simple reason that with those large pipes (probably 3" or larger), the water will be moving very slowly.

I'd use a Taco 0010 that will give you the minimum 25 gpm, 3.5' head. If you take a look at the performance curve for that circulator, it's nice and flat.

Of course, you can always experiment and try the next size down (Taco 007) and see how it works.

BTW, can you share with us the method you use to show the delta sign?

This is a thread from a previous discussion on pump sizing (very long):


Date: December 19, 2002 09:01 AM
Author: Boilerpro (boilerpros@cin.net)
Subject: Sizing Circulators

Continue to do alot of gravity conversions and sizing these pumps is again on my mind. This is the info I have so far:

Gravity systems usually were designed for a 30F to 40F delta tee from boiler supply to boiler return.... 180F out 140F to 150F return.

Most modern systems were designed for a 20F delta tee.

The radiation on many gravity systems is greatly oversized.... largely due to insulation added to wall/ ceilings, storms added, etc.

Steamhead's essay at Hot Tech topics says to base the size of the circ. on gravity conversion on radiation size and use a circ 50% Larger than that of a modern system. I.E. 500 edr installed radiation times 150 btu/edr equals 75,000 btu/hr. In a modern 20F delta tee system this would be about 7.5 gpm....now we need to add another 50% which gives us about 11 gpm.

However, many also say to try to not disturb the operation of the gravity system when installing a new boiler and pump to achieve the same flow balance To me this means we want the same flow rate through the system. In other words we need to size the system pump for the added pressure drop of our new smaller piping while maintaining the same temperature drop across the system (assuming the system still needs to put out the same amount of heat and run at 180F.... which most don't in my experience). Using the above example: 75,000 btu/hr installed edr at, let's say, 35F delta tee.

75,000 btu/hr divided by (8.3 x60 x 35F delta T) equals only 4.3 gpm.

Sooooooo, Which is right 11 gpm or 4.3 gpm?

Boilerpro


TopPreviousNextPrintReply

Date: December 19, 2002 10:58 AM
Author: Mark J Strawcutter (mjstraw+wall@iup.edu)
Subject: I seem to remember

a posting from Steamhead where he says he's ended up with a Taco 005 on his own home's system (gravity conversion).

This would seem to indicate that 4.3gpm is the right answer.

But it would also seem he's contradicting himself :-)

Mark


TopPreviousNextPrintReply

Date: December 19, 2002 03:35 PM
Author: Mark J Strawcutter (mjstraw+wall@iup.edu)
Subject: no, he's not

Oops - was looking at the wrong pump curve.

an 005 will give you 11gpm at 6-7ft of head (which is probably a bit on the high side for head required on a gravity conversion).

Mark


TopPreviousNextPrintReply

Date: December 19, 2002 12:00 PM
Author: Mike T. Swampeast MO (mtman@clas.net)
Subject: Delta t & Gravity Systems

Glad someone mentioned the HIGH delta t that many gravity systems were designed around.

It ties into my ideas of why the rads are so big to begin with.

A good post-WWII heating book shows equations for computing the flow in gravity systems--they're quite hairy. The goal is a very narrow specific velocity range in the pipes. You then computed the actual amount of flow based on this velocity and ensured it was adequate to serve the radiation attached from that point on. Process repeats throughout the system--but since changing one thing later affects everything before it gets strange. To make it even more complex changing the supply temperature changes the delta t which changes the velocity which changes the flow...argh!

The good thing is that the "old way" of sizing the boiler/pipes/radiation where you choose a boiler based on supplied radiation; initial main size based on total output and subsequent main size based on cross-section of rad valves still to be served usually came pretty close to the difficult calculations. I'm not even certain that the equations existed back then anyway...

Since the motive force in a gravity system is the difference in density (temperature) of the water between the supply and return, you can actually view the radiation as being the circulator. Since it's the prime place where heat is extracted, it really is providing the motive power for circulation. Consequently, the rads have to be large enough not only to heat the room, but to extract enough heat from the supply to create circulation.

Try to extract too much heat (too large a rad or too small a pipe) and the delta t increases to a point so high that the velocity it requires to maintain the delta t overwhelms the friction loss of the piping.

Try to extract too little heat (too small a rad or too large a pipe) and the delta t decreases so much that you loose the motive force for driving the system in the first place.

The real balancing act came from sizing the pipes and rads so that the VELOCITY remained nearly constant through the entire system.

When you consider multi-floor applications where there are different elevations on the rads (thus a different velocity for an otherwise idential rad/pipe combination) it becomes a near wonder that these ever worked to begin with. You'll understand why those "restrictor plates" are sometimes there--generally when the "rule of thumb" sizing methods produced a very different result than the actual calculations would have produced.

Remember also that as long as that coal fire burned, these systems produced heat. But it wasn't too practical (or even possible) to keep the fire "just right" at all times. They needed the ability to really "crank" the heat to compensate for the fire nearly dying on a cold night--thus that 180 degree supply temp ability. If you remember that tendency to "air the house" each day, they REALLY needed to be able to crank the heat...

Go back again to VELOCITY. It was SLOW. The BTUs "hung out" in the radiator for quite a while and many of them were able to find their way to the room--thus the hefty delta t AT DESIGN TEMP. Since the boiler (ideally) was always supplying heat, the rads didn't go through hot & cold cycles--they just got warmer as the fire was made bigger as it got colder outside. If everything was sized perfectly, the velocity would increase proportionally to the size of the fire/temperature of the water BUT the velocity would stay the same through the system as a whole at any given supply/return temp combination.

NOW, throw in a digitally controlled heat source (the boiler with thermostat) and a circulator.

Velocity no longer changes with supply temperature. Supply temperature varies GREATLY and the rads go through "hot and cold" cycles. It's now behaving much more like a steam system--just at lower temperature--and the mass of the radiators keeps the space temperature from varying too much.

Delta t no longer has slow, gentle changes as it did when operating under gravity. When the circulator is off the water in the rads cools greatly. When the circulator kicks in, delta t is huge--thus the need for protection from low return temps. BUT velocity is much higher--and constant--than under gravity and the BTUs don't "hang out" in the rads as long--thus fewer make their way to the room. It's the added velocity alone that makes enough BTUs available to heat the space.

As the weather cools outside the circulator runs more frequently and there is less on-off difference in temperature of the water in the radiators. Delta t actually decreases! This is the EXACT opposite of what happens in a system with gravity flow. Fewer and fewer BTUs make their way from the water to the space with each pass through the radiator. You have to keep the velocity (flow) up to ensure enough BTUs can make their way to highest, farthest rad. This rad will have the highest delta t (lowest velocity) in the system.

FINALLY:

Sooooooo, Which is right 11 gpm or 4.3 gpm?

The 11 gpm is the most correct figure in most every instance:

With simple control (on-off circulator): You need this amount of flow so that you put enough BTUs in the lowest, closest rad OFTEN ENOUGH to liberate enough heat at design. This rad will have the lowest delta t (highest velocity) in the system. Remember that flow restriction is virtually non-existant in these systems thus the "downsizing" of the mains at the boiler and the high flow/low head circulator. As long as the velocity is not so high as to prevent any BTUs from getting off in the rads, you won't have a problem. If this only happens on an occasional rad (likely ones closest to the boiler) you can "fix" it by going back to the restrictor plates. Remember that you likely have a bypass line installed around the boiler anyway so a good chunk of flow never makes it to the rads to begin with.

If you get the idea that you could eliminate the bypass by using a smaller circulator--thus trying to make it flow the way it did under gravity--watch out. You'll wind up with extraordinarily high delta t--particularly during mild weather.

If you now think "I'm using a condensing boiler so I don't have to worry about return temp" think again. That careful balance of velocity in the gravity system DOES NOT WORK when you add forced circulation. Instead of the radiators "pulling" the water through the boiler, the circulator is now "pushing" the water through the rads. Since the circulator is now the only source of this motive power, its energy is expended in the EASIEST way possible--the path of least resistance--NOT the path through all the radiators. While every molecule of water may be moving simultaneously, they WON'T be moving at the same velocity as they did under gravity. About your ONLY prayer of getting this to work would be with a single-floor reverse-return piping arrangement.

With constant circulation, TRVs, normal boiler: You wind up with the problem just mentioned--incredible delta t. You also have to build enough head pressure to keep the TRVs functioning so you still need the ability for high flow--you just need even more bypass around the boiler much of the time.

With a proportionally fired condensing boiler, constant circulation and TRVs things can change. While they won't have large enough tappings to operate under gravity circulation, this should be possible: If you could somehow vary the velocity of water directly with the burner, your flow could be reduced immensely most of the time. The TRVs will even the velocity for you by adding restriction as needed. I have no idea though if the reduced electric consumption would be worth the much greater complexity--I sort of doubt it...


TopPreviousNextPrintReply

Date: December 19, 2002 04:03 PM
Author: Boilerpro (boilerpros@cin.net)
Subject: Verrry Interesting...

You brought up a good point about the delta tee changing with the load on the system. However. I'm not so sure about the idea that the delta tee is huge in warmer weather on a pumped system. This does seem to be the case with a strictly gravity system, though. The sonstant and high flow of cold water going into the heating plant keeps the supply really cool. If you are running 11 gpm at 20F delta tee, you need a 110,000 btu/hr to maintain this temp difference. If you are running a 60F delta tee at 11 gpm, you need a 330,000 btu/hr to maintain this delta tee. Using the 500 edr load, you would only have about 80,000 btu/hr available (assuming th rare instatnce wher the boiler is sized properly). I don't see any way you can get the delta tee up higher unless you slow the pump. Also, Mike, you said the converted system would act more like a steam system, very large temp fluctuations in the rads. I just haven't found this to be true. The sheer mass of iron, steel and especially water in these systems prevents that. On a system I converted a few years ago, where the heat load was very close to the radiation capacity and the boiler was sized to the heat load, the radiators typically only changed about 20F on a typical (20F) winter day from the start of the 20 minute long firing cycle to the end. Hadn't thought about the variation in flow through a gravity system at canging water temps... wa thinking only at design. Thanks for bringing me back to the fact that all systems are dynamic, forget that for a moment!

Boilerpro


TopPreviousNextPrintReply

Date: December 19, 2002 06:03 PM
Author: Mike T. Swampeast MO (mtman@clas.net)
Subject: YOU'RE RIGHT

I exaggerated and may well have goofed the delta t difference in some cases. Most of my good measurements have come AFTER my TRVs and bypass were installed and that tends to really skew the results. AND most all of the measurements have been on my system with the double-size (at best) boiler and radiation that varies from 2x oversized to nearly 4x (only one fortunately and most is about 2½ times).

I was also looking at the gravity system as one in constant operation by virtue of the constant fire--still operating with solid fuel--not considering the "cold start" when delta t WOULD be UTTERLY massive.

Oppositely I was considering the type of "conversion" I see most--older cast iron boiler of high water content when the boiler always heats to high limit (RARELY reset) and the circulator is controlled by the t-stat. In that case, the delta t really does act as I described--highest in moderate weather dropping as it gets colder outside.

If you consider a much more modern boiler installation though it has LOTS of bypass in order to keep the return temp AT THE BOILER above 140 degrees. BUT the return temp from the system is MUCH lower. This is my goof, because I was considering the delta t as the difference in the temp AT THE BOILER with the difference in the return BEFORE the bypass. I probably should have kept both references on the one side or another of the bypass. Or should I? Am not certain.

"On a system I converted a few years ago, where the heat load was very close to the radiation capacity and the boiler was sized to the heat load, the radiators typically only changed about 20F on a typical (20F) winter day from the start of the 20 minute long firing cycle to the end."

I'M NOT TRYING TO BE DIFFICULT AT ALL BY SAYING THIS BUT if the system were still operating with a perfect coal fire (or other continuous proportional fire) the temp of the rad wouldn't have changed at all and the delta t would remain constant. If you put that 20° in terms of the assumed delta t of 20° it gets pretty big...

"I don't see any way you can get the delta tee up higher unless you slow the pump." You CAN'T as long as the fire is EXACTLY SUITED TO THE HEAT LOSS. With an on-off fire though you can get a higher delta t given the same velocity (pump speed). It happens only when you raise the temperature of the space itself--it then decreases somewhat as the temperature of the space cools. (At least I think that's right.) When raising the temp you increase convection which speeds heat transfer. Think of the whopping delta t you can get across a window rad... In this way a proportional burner is demonstrably more efficient as it strives only to MAINTAIN temperature--not change it. TRVs do this as well, but only at the radiator--not the fire itself.

Am currently working on reading (and understanding as best I can) the book "Maxwell's Demon" that someone recommended here. GREAT stuff. That impossible demon explains why it would take some reduced (and finite) amount of energy to exactly maintain the temp of a space than to maintain the exact same average temperature...


TopPreviousNextPrintReply

Date: December 19, 2002 09:47 PM
Author: Mike T. Swampeast MO (mtman@clas.net)
Subject: FINALLY found the BIG problem...

...that made you say, "how can you increase delta t without decreasing flow?" That previous WASN'T what you were talking about!!!!!

"You have to keep the velocity (flow) up to ensure enough BTUs can make their way to the room."

Should be "You have to keep the velocity (flow) up to ensure enough BTUs can make their way to the HIGHEST, FARTHEST rad." ...this radiator will have the highest delta t (lowest velocity) in the FORCED system.

AND "You need this amount of flow so that you can put enough BTUs in the rads often enough to liberate enough heat at design."

Should be "You need this amount of flow so that you can put enough BTUs in the LOWEST CLOSEST rad often enough to liberate enough heat at design." ...this rad will have the lowest delta t (highest velocity) in the FORCED system.

SORRY!!!! I FORGOT THE DYNAMICS MYSELF AT THAT POINT!! You ALWAYS have to remember that many of the dynamics in a gravity system under natural flow are ESSENTIALLY OPPOSITE when you introduce forced flow.


TopPreviousNextPrintReply

Date: December 19, 2002 09:55 PM
Author: Boilerpro (boilerpros@cin.net)
Subject: Got to admit it does get confusing....

I think I am going to need to ponder these things some more...starting to get cross eyed. Hope Steamhead will weight in on this topic too.

Boilerpro


TopPreviousNextPrintReply

Date: December 19, 2002 02:24 PM
Author: Duncan (duncanw@mindspring.com)
Subject: No numbers, but an interesting experience.

I saw a 50 room, 110 year old, three-story apartment building that had been converted from one pipe steam to two pipe hot water. There were three inch mains in the crawl space, one to one and a half inch risers to rads as I recall. Urea-formaldehyde foam insulation had been pumped into the building.

All of the rads - at least forty or so I think - had 1/2 inch copper returns installed.

The water in this large system was circulated with a Grundfos UP-2664. In other words, you don't need much pump.

If it were me, I'd size the circ based on the longest run. I'd install the smaller circ between isolating flanges. If it turned out wrong, I'd switch to another pump. It's really surprising how little pump you need most of the time.

By the way, the building owner (at the time) who did the changeover of this big old system? He was an engineer who specced circulators for commercial jobs. The building is in Cape May, NJ, 55 Jackson Street, formerly known as the Rivendale.


TopPreviousNextPrintReply

Date: December 19, 2002 06:05 PM
Author: Boilerpro (boilerpros@cin.net)
Subject: Those 1/2 inch returns really help

They prevent the big "short circuits" that occur with regular gravity conversions. I did the same thing, but with 1/8 inch pipe nipples for returns installed where the steam air vents were on the rads.

Boilerpro

Attachments:



TopPreviousNextPrintReply

Date: December 19, 2002 09:56 PM
Author: Mike T. Swampeast MO (mtman@clas.net)
Subject: EXACTLY

I REALLY liked that idea when you presented it a few months ago.

But that's a steam conversion...not a gravity conversion of course.

Something tells me that if you wanted to put a restrictor plate (with just the right size hole) in EVERY rad on a gravity conversion that you COULD decrease the flow rate considerably...

Talk about going cross-eyed!


TopPreviousNextPrintReply

Date: December 19, 2002 10:27 AM
Author: Greg (gregolma@olm1.com)
Subject: pump sizes

I'm not an expert, but I have been conducting a little experiment on my own converted gravity system.

A 3/4" B&G 100 or the like is sort of the norm for a system pump on a gravity conversion. I used that as well.

But I was concerned that some rads were not as warm. I have about 1,100 sq f of rads in a house nearly 4,000 sq f. I wondered how much a bigger pump would affect the system, inspite of the fact I had read all the info.

Since it was my own home I had the luxury of time. I happened to come upon several old B&G pumps sized from 1" up to 2". I installed the 2" and found that air in the system no longer ended up at the bleeder, but instread on th3e upper floors radiators. Too much flow.

So I have continued to switch out the pumps and now I'm at 1-1/4" and I get a little air at bleeder and some upstairs.

The next step is to swap in the 1".

So it seems that both practice and theory are in synch.


TopPreviousNextPrintReply

Date: December 19, 2002 04:14 PM
Author: Paul Pollets (Paul@advancedradiant.com)
Subject: Gravity Conversions and pumps

I prefer water lubricated circ pumps. I've used a 15-42 Grundfos for small systems 26-64 for larger and have used a 26-99 for one conversion with 30 radiators and 4" mains. It's wise to put TRV's on the radiators to balance the system and install a pressure bypass differential (Danfoss AVDO) around the pump. The PBD will also reduce the pumps head if it's too large. Leaving the radiators uncontrolled after the conversion usually results in callbacks and balance complaints.

To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"


TopPreviousNextPrintReply

Date: December 19, 2002 09:39 PM
Author: Steamhead (steam.head@verizon.net)
Subject: I actually get about 15 GPM

from the 005. That's a bit more than the calculations on which the chart is based would indicate but it still works much better than the 30-GPM B&G 100 did.

That "50% bigger" rule came from B&G's 1940 Handbook. The actual method used in the Handbook was to calculate the circ size and then use the next size larger circ. Based on the circs B&G was making at that time, the "next size" would give about 50% more capacity, give or take. Taco had a similar procedure. The reason I express it that way in the chart is because B&G and Taco aren't the only ones making circs these days, so this allows you to go straight to the performance curves of whoever's circs you're using.

BTW, everything I've read on gravity conversions says to choose the circ's capacity at a 3-1/2-foot head. This is because there's so little resistance in old gravity pipes. Most of the resistance on a conversion is probably in the near-boiler piping!

Swampeast Mike T. hit it right on the head with his Delta-T post. This is what got me started down the circ-sizing path. If the water is moving too fast, the heat is not being picked up properly in the boiler or released in the rads. This will show up as a very small Delta-T. We want to try to mimic the flow that would occur with a gravity boiler at 180 degrees.

In any case, I would not have sent the chart to Dan without having tried it myself on several different-sized systems. It worked so well at my house that I was able to slightly down-fire my boiler!

To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"


TopPreviousNextPrintReply

Date: December 19, 2002 10:49 PM
Author: Boilerpro (boilerpros@cin.net)
Subject: Say, Steamhead

Here's the issue that is confusing me. A gravity boiler at 180 degrees is probably running about 30 to 40 delta T. About double that of a modern system. This means that the flow rate through the gravity system is approaching ONLY 1/2 THE FLOW OF A MODERN 20F DELTA TEE SYSTEM. However, by sizing the circ to 50% bigger than a new system WE ARE REDUCING THE DELTA TEE TO ONLY 15, NOT INCREASING IT TO 30 OR 40. A bigger circ is going to make the system run less like it originally was, so we are not mimicing the original flow at all. Am I off my rocker?

Boilerpro


TopPreviousNextPrintReply

Date: December 20, 2002 02:14 PM
Author: Steamhead (steam.head@verizon.net)
Subject: The difference is

you have all that water in the large pipes of a gravity system. If you size the circ strictly to the BTU load or EDR, the system will be sluggish. You have to move the water fast enough, but not too fast.

To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"


TopPreviousNextPrintReply

Date: December 20, 2002 06:13 PM
Author: Boilerpro (boilerpros@cin.net)
Subject: Sluggish

Decribes the operation of all the gravity systms I've seen, both pumped and not pumped. So we really aren't trying to mimic the original operation of a gravity system. Hope to get a chance to toy with this on some system, maybe my neighbor will let me. Steamhead, have you tried to run your system at the 4 gpm level? I would imagine it would respond only slightly slower because it has to warm up all that water in the supply pipes first, but once going, a btu is a btu delivered at the rad, whether running only 15 delta T with 11 gpm or 40 delta tee at 4 gpm. I do agree that there are alot of overpumped coverted graivty systems. I've pulled out a few 1/6 HP 2 inch pumps in put in NRF-22's. Just wondering if we can go back to the orignal flow rates.....more or less in view of what Mike posted about the variable lfow rates these systems probably used to have.

Boilerpro


TopPreviousNextPrintReply

Date: December 20, 2002 07:01 PM
Author: Steamhead (steam.head@verizon.net)
Subject: Interesting idea

but I'd need to locate a circ with a lower delivery rate than the 005, that would fit standard mounting flanges. Part of the "test-bed" idea is easily swapping components in and out so I can see the effect, and change it back quickly if it doesn't work as well. The 006 has a lower rate as does the 003, but they have sweat connections and I'd have to do some repiping. Maybe when I get some time....

Most of the gravity systems I've seen really got going as the boiler temp reached 160 or so. Trouble is, on mild days they don't get nearly that warm and circulation suffers. If we can provide that 160-or-over rate of circulation all the time no matter how hot the boiler is, we should be fine. I think it's OK if we go slightly over, but not so much as to get a diminished delta-T.

To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"


TopPreviousNextPrintReply

Date: December 20, 2002 07:40 PM
Author: Boilerpro (boilerpros@cin.net)
Subject: Why not just add a flow meter

to your existing pump and throttle it with a valve? No need to change the pump, until you find out where you wnat to be. Just thinking out loud.

Boilerpro


TopPreviousNextPrintReply

Date: December 20, 2002 08:24 PM
Author: Steamhead (steam.head@verizon.net)
Subject: Good idea

but I've never used one- any out there that you like?

To Learn More About This Contractor, Click Here to Visit Their Ad in "Find A Contractor"


TopPreviousNextPrintReply

Date: December 20, 2002 09:13 PM
Author: Boilerpro (boilerpros@cin.net)
Subject: Haven't used any yet

But I think there's plenty of other folks around here that can steer you in the right direction.

Boilerpro


TopPreviousNextPrintReply

Date: December 20, 2002 12:15 AM
Author: Nobby (nobbypeers@earthlink.net)

Hmm very interesting enjoyable discussion and a topic that has been on my mind as of late. My 1900's house has as a very poorly converted gravity system still using the original piping new boiler, with items such as woefully inadequate expansion tank (4 gallons, being changed next week), circulator on return side to list 2 issues.

Well circulator size was my recent consideration. Right now I am running on a single Taco 007 which seems fine and matches Steamheads table. Balancing was won in the following way. Firstly I split the house into two zones. The old system has two return/feeds which split the house down the middle equally conveniently North/South. I placed the thermostats downstairs accordingly. I then proceeded after repacking the rad. valves (which hadn't moved in years) to balance the system by adjusting the valves. Now I fully appreciate that this is very possibly not a contractor solution due to the need for a period of 'fine tuning'/callback. The zoneing would ease balancing by reducing the amounts to deal with in by half(and also make my cold North Side warmer:). I rough balanced the system by letting the system cool of completely and then one zone at a time run it and run around the house like a blue-arsed fly constantly feeling feed pipe temperatures and adjusting valves accordingly. I also marked the valves to aid in this.

The interesting thing is that I was suprised in that the rough balancing was fairly close to the final result. Some tweaking has since occured but I am finding the need to resist fiddleing to much. I should also add that it still behaves balanced when both zones are on, I was lucky in that I was able to split the system fairly equally with fairly comparable lengths of run, # of rad. etc. I am replumbing the kitchen and am piping this with two pipe reverse return(new zone) using old rads, to see how much this self balances. I should add that my profession is that of a Marine Engineer on older vessels so that constantly monitoring and adjusting various fluid piping systems is part of the deal for me. I even kept a basic log of my Rad. Valve adjustments. Now I appreciate that my fine tuning is over the top but I am curious as to the validity for a contractor to rough tune a system by allowing complete cool down as aforementioned? The thing is it is not a simple case of saying to the HO oh just turn the offending rad. up or down if needed as it will have a knock-on effect. Plus the danger of inadvertantly turning valves off. Also the valve turning is pretty much a matter of a couple of degrees sometimes to make a difference. Now TRV's nice idea cannot bear the thought right now of breaking apart all that piping though. One ship I worked on had a gravity system running on an old furnace (anyone heard of PYRO Scandinavian marine boiler firm?) that used the fuel bowl carburetor forced draft setup. The furnace was always on with very little automation i.e I usually turned up or down the carburetor manually to adjust water temp. setting accordingly. The common complaint though was on the nights when the weather improved someone would turn their rad. down the next thing I knew was a very hot and bothered crew member from the next cabin whose cabin had become quite the mouldy old sauna.

Is there a science to the use of restriction plates or is it trial and error seems to me a PITA to have to change them out if you don't get it right.

My latest plan is Outdoor Reset when the next paycheck rolls in.


TopPreviousNextPrintReply

Date: December 20, 2002 08:24 AM
Author: Boilerpro (boilerpros@cin.net)
Subject: Adjusting valves....

It's great when you can turn them! When you can, you can probably get virtually any pump to work. If you really want to see how far you can go and save a little on you electric bill, you may want to try downsizing the pump some more and allowing the delta tee to increase. with those great big pipes it only takes a few watts of power to move that water. You may want to try for that 4 gpm instead of that 11 gpm in my example. BE curious to see how its works. Can't mess with my own converted gravity system, because it no longer is....I've repiped it with new smaller copper piping during ou renovation.

Boilerpro


TopPreviousNextPrintReply

Date: December 21, 2002 09:11 AM
Author: Boilerpro (boilerpros@cin.net)
Subject: "Oversized boilers" in Gravity systems

Mark's and Steamhead's comments about flow being very slow until the system got warmed up got me thinking. Maybe this is why so many old time contractors insist on sizing boilers to the radiation.....it gets the flow moving faster if the boiler temp rises quickly. They later added pumps to thier new boilers, but forget the reason why they were making those boilers so big was for circulation, not heat. Maybe another great mystery solved.

Boilerpro

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

R. Kalia

wow, a lot to read

I confess I haven't figured out all the stuff in those posts yet, but I'll keep re-reading them.

You hit the nail on the head with your suggestion of the Taco 0100. That's what we currently have---or rather, a B&G series 100, which has much the same curve. It's been there since the early 1970's (we bought this house last year). Our contractor wants to cut the flow by a factor of four to get δT=20. As you suggested, maybe we'll split the difference by getting a Taco 007 or B&G NRF-22.

You can copy the upper-lower case deltas from these posts and save them in a document somewhere, or (in Windows) you can get them from the Character Map. In my computer, the character map is under Programs | Accesories | SystemTools, but this varies. Use the Courier New font to get characters from lots of different scripts.

Ken_8

That's The record folks!

The longest post in the history of the Wall!

Thanks to MS "copy and paste"

I can't even read my optomitrist's card now to make the appointment...

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

Mike T., Swampeast MO

Gravity Flow

Gravity systems [seem] to have been designed around a 30°-40° delta-t in the radiation at maximum boiler output.

To get an idea of the original flow in your system:

160,000 btu/hr (from the radiation) / 30 (delta-t) = 5,333 pounds of water per hour. (Pounds are easy for this because of the definition of a btu and measure of btu/hr. It gets kind of hairy when you need to convert to gallons per minute--something I don't believe the dead men were concerned with...)

Assuming 180° water entering the radiation and the 30° delta-t for an average of 165°, those 5,333 pounds of water occupy about 87.57 cubic feet (0.01642 cubic feet per pound @ 165°).

87.57 cubic feet = 151,321 cubic inches.

One gallon = 231 cubic inches so, flow is about 655.1 gallons per hour or 10.92 gallons per minute.

-------------------------------------------------

I'm guessing that your 160,000 btu/hr estimate is coming from the output of the radiation at 20° delta-t with 180° water. As you know, your system probably NEVER needs to put out anywhere near this amount of heat and you should be sizing a replacement boiler to the actual heat loss of the house, not the ability of the radiation to put out heat.

-------------------------------------------------

The problem with forced circulation of a gravity system is that under gravity elevation above the boiler increases the motive force for circulation. In a multi-floor system the piping had to be designed to retard flow to the upper floors while encouraging it at the lower. With a circulator the water now wants to "short cut" through those low and very open passages. With almost no restriction in the piping it seems that you have to move water much faster than necessary to ensure flow through the entire system.

----------------------------------------------------

When you size the boiler to the heat loss you get a strange situation. If connected straight into the system with a high-volume, low-head circulator you wind up with extremely low delta-t due to the high flow, and extremely low supply temperature due to the output ability of the radiation, but you [seem] to need the high flow to guarantee full circulation...

I believe this is why condensing or condensing/modulating boilers invariably use some form of primary-secondary piping. Your flow through the system itself stays high and its delta-t low, but the boiler loop will have much lower flow and the higher delta-t required to keep the efficiency of the boiler at a high level.

If you add TRVs on all of the radiators it [seems] to get even more bizarre. Delta-t in the radiation now goes up--WAY up--and flow goes down--WAY down. But this mainly happens when the system is maintaining temp--when heating (particularly from cold start) flow is still high and delta-t almost imperceptible.

R. Kalia

Your calculation is the same as mine, give ot take 10%.

My question was, what about the URL given in the first post? It seems to say (and Alan Forbes also seems to say) that you need higher flow than calculated for the required heat loss and a 20F delta-T.

You also seem to be saying that the higher flow is required for the boiler, not for the converted gravity piping itself. But that's not what the article in the original URL is saying. In fact the article says the flow should be determined by looking up the table for the original design BTU, not the (lower) actual BTU.
.

Mike T., Swampeast MO

The high flow seems to be a requirement of the piping, not the boiler. In fact I don't believe that any boiler will be operating at peak efficiency when delivering say 80 mbh @ 15 gpm (or even higher).

Gravity systems don't break laws (in fact they follow the simplest of laws to the "tee") but it seems to be difficult to apply laws (in familar modern terms) to these systems after a circulator is added.

Gravity systems had VERY little motive force to work with. A 30° delta-t gives about 122 milinches of water column pressure (or about 0.0044 psi) for each foot the radiation is elevated above the boiler. Restriction in the piping to each and every radiator HAS to be below this amount for circulation to occur.

With the general assumption of 10' per story, first floor rads will have about 1,220 miliinches (1.22 inches) of water column pressure--consequently the main, branch, fittings, valve of both supply and return must offer less than 1.22 inches of restriction.

At the 2nd floor you get double this amount of force or about 2.44 inches of water column; 3rd floor you get thrice or about 3.66 inches of water column.

Were this being designed for forced circulation you'd use your assumed delta-t, heat loss and the MAXIMUM head loss in the system to size your circulation. But, gravity doesn't work that way... You essentially have to consider each and every radiator to be its own heating system. When the radiators are on different floors, they might as well be in different structures considering the multi-time relative differences in pressure. Were this a modern system with such drastic relative differences, you'd be going back to the drawing board if you wanted to use a single circulator...

Getting the gravity (particularly multi-floor) systems reasonably balanced is not much shy of an engineering marvel. Branches feeding upper radiators would be intentionally more restrictive and those lower intentionally less restrictive as the goal is to balance velocity of water through the entire system. Only when the velocity is balanced can you expect your delta-t in each radiator to be reasonably balanced as well. Only when the delta-t is reasonably balanced will your actual radiator output be reasonably within your output calculation. Throw any one off and everything goes kerflewie!

If your system was designed for about 10gpm originally under gravity circulation with about 160,000 btu potential in the radiation, you likely have two pairs of mains at least 2½"--and likely 3" if there is much in the way of horizontal distance. By modern sizing standards those mains are easily capable of handling 170-260 gpm of flow! With a 20° delta-t that means they're capable of transmitting 1,700,000 - 2,600,000 btu/hr! Think elementary school!

Install a circulator and a boiler of comparatively tiny volume (and likely smaller heating capacity as well) and everything does go kerflewie! One, I don't think it's particularly easy to design a circulator to move only 4-5 gallons per minute with a head pressure of less than 4 INCHES through a system that holds hundreds of gallons of water. Even if you could, that old gravity sizing is going to bite you in the **** big time and those 4-5 gallons per minute will find the "easy" path through the lowest, closest radiators.

So, you have to move much more water than necessary to increase friction to movement and build head pressure for the circulation. While the differences in restriction in the piping between rads on diffferent floors are relatively large, they are tiny in absolute measure. Build restriction through the system as a whole and the relative differences between the radiators and that of circulator diminish and you again have reasonably balanced flow.

When it comes to the boiler itself, again this doesn't seem to be a good condition for any boiler in terms of efficiency. A number of "solutions". A significant bypass of supply water right back to the boiler return will allow the boiler to come up to its preferred operating temperature and [hopefully] burn for a reasonable length of time in moderate weather--problem however is that this hasn't changed the total flow through the boiler and delta-t at the boiler does a nosedive with a somehow corresponding drop in efficiency.

Primary-secondary is probably a much better solution than simple bypass.

Image below is from the Viessman Vitodens so-called "low-loss header" that is required on gravity conversion systems. (It really is just a primary/secondary connection but with a bleeder, "dirt collector" and sensor.)

Pay careful attention to what it's doing (or is that assuming?):

1) It keeps flow through the boiler LOWER than flow through the system.

2) It keep the boiler supply temperature HIGHER than system supply temperature.

3) It keeps boiler return temperature EQUAL to system return temperature.

4) It keeps boiler delta-t HIGHER than system delta-t.

If you're installing a new boiler on a gravity conversion system and you don't have TRVs, I would highly suggest that you make (or have made) something similar. Just ENSURE that your flow through the boiler side remains significantly LOWER than the flow through the system and you should get a similar result even without the temp sensing and modulating circulator in the Vitodens.

With TRVs? PROBLEM! Put TRVs on a gravity conversion system and delta-t across radiators returns to (or even exceeds) the original delta-t because you have added a LOT of resistance to the system. Consequently flow drops back to "normal" levels--possibly even lower considering the high delta-t. In that case, you have a REALLY strange situation...

R. Kalia

I'm, like, totally confused

In our current system (design ~150MBU, actaul needed ~80MBU), we have a series 100 moving maybe 25 gpm through the former gravity piping. In contradiction to what I have read elsewhere, the 2nd/3rd floors are too hot, not too cold, with the current forced circulation system. Maybe its the new attic insulation, maybe the orifices were removed, or maybe we are pumping too hard.

To fix the overheating problem on the 2nd/3rd floors we had TRVs installed there (the thermostat is on the 1st floor, and no TRVs there). This works well now.

So my guess is, the pumping can be cut back some--flow to the upper floors will be weaker, so the TRVs will open up to hold temperature, but that is fine. The question is, how much lower flow is OK? half? One quarter? No way to know. I guess if the 2nd & 3rd floors stay cold, it tells me the new pump is too weak. Nothing like experimental science.

Steamhead

If you have radiation that can emit 150 MBH

that would equal 1000 square feet EDR, and based on that I'd say your current circ is the right size.

You have to remember that even if you have TRVs, you still have to be able to supply all the radiators with water at the same time. There will be times when all the TRVs are open, and a circ that's too small will not supply enough water to heat all your radiators. An example of this would be when the system comes out of a setback period, and the lowered temperature has caused all the TRVs to open.

The reason you saw the old B&G and Taco manuals saying to increase the circ size on a gravity conversion is because the pipes were so big, and held so much more water than those in a system designed for forced circulation. This overcomes the sluggish response caused by all that water. But all too often, circs on old gravity systems are oversized way beyond that 50%, and this can actually impair the transfer of heat from the boiler to the radiators.

Another thing that might have caused the second and third floors to overheat is oversized radiators. If this is true, they were probably oversized to accomodate people who liked to sleep with open windows. Or, radiation might have been removed from the first floor. But you sound like you have this problem nailed.

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

Mr Takagi

I agree with PDQ!

It seems like there is no real answer to circulator size on these gravity conversions. I had started a thread on choosing between a Taco 007 or a B&G100 for the same type system, but this one is much more interesting In my situation I am putting in a Weil Mclain Ultra 155 to replace my old boiler. I will have it piped P/S with the Taco 011 as the boiler circulator, and who knows what for the system pump. The old system uses the B&G 100 and all the rooms heat evenly, so is it safe to assume that it is the correct size? Or would a smaller pump be more efficient. I calculated the output of my cast iron rads. at 110MBU/hr and total system volume is 150 gallons of water. According to Steamheads chart, I only need 15gpm so the B&G100 is oversized. I may just use the 100 in the new system, and throttle back the flow using one of the isolation valves on the system loop to see how it works with less flow. Does this seem logical?

Mike

R. Kalia

bang-bang vs. continuous circ

> The reason you saw the old B&G and Taco

> manuals saying to increase the circ size on a

> gravity conversion is because the pipes were so

> big, and held so much more water than those in a

> system designed for forced circulation. This

> overcomes the sluggish response caused by all

> that water.


Am I right to assume that sluggish response is a problem on 'bang-bang' systems (like mine is), but not on continuously circulating systems (like mine is going to be)? I mean, our system currently takes forever to get warm after a call for heat, but then it overshoots. This is not an issue wih continuous circ.

So maybe lower flow is OK? Reason I'm pushing this so much is the delta-T issue, we'll have a condensing boiler and there's some loss of efficiency associated with the small delta-T that comes with high flow. Plus a bigger circulator takes a bit more electric power.

R. Kalia

Consider the Grundfos 15-58, as someone else recommended in the other thread. Apparently there is not a significant cost difference. You can play with the flow rate and find out what works best.

Of course even the fastest flow rate of the 15-58 will be smaller than that of the B&G 100; it is about the same as the Taco 007. But you, like me, are upgrading to outdoor reset and thus almost-constant circulation, and I claim that the flow rates in the original article are not necessary since 'response' is more of an issue with start-stop heating using 180F water. Then there is the other issue, i.e. whether you'll be pumping so slowly that water will take only the easy route and avoid the longer routes via the second floor. I don't have an answer for that.

PS If your radiators can output 110MBU, your house must need less than that. So why are you oversizing your Ultra boiler so much? True, it modulates, but the bigger boiler has a higher low limit than a smaller boiler, so a better-matched boiler (the 105 or the 90) would be more efficient as well as cheaper. There is no advantage at all in having a boiler that puts out more than can be dissipated...a chain is only as strong as its weakest link. Heat loss calculations tend to be quite conservative, i.e. they give higher numbers than the actual heat loss.

Steamhead

On paper it looks like that

but with big cast-iron radiators that hold their heat, you won't see big temperature fluctuations.

BTW, how did you come up with your BTU figure? Did you convert it from the rads' EDR ratings?

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"

Mr Takagi

Why I'm going with the Ultra 155

My IBR calculations come out to between 85-88MBU. Allowing myself even 10% room for error makes it more like 95MBU. Since the IBR on the 105 is only 81MBU, I figure it's safe insurance to go with the larger 155 in case my numbers are off at all. Plus if I add DHW down the road I'll have plenty of output for that. I'd rather spend a few hundred bucks extra now and know that it will do what I want this winter. It's still going to be a huge improvement over the current 210,000 input boiler!

Thanks for all the help. Once again the wall has proven itself to be a huge wealth of information.

Mike

Weezbo

I read your Out line...about \"pumps\"

i am P Poor communicator..had i read what you wrote ..i wouldnt have repeated what you had already stated.
This isnt exactly related however...have you installed any of the new i-series zone valves from Taco? Alaska is way off the beaten path so i just got a copule few yesterday and went and plumbed them in...this will be my first install using the outdoor reset on the radiant Staple upwith heavey aluminum plates,and a set point for a bypass, on an oil fired boiler....my question is have you found them to be reliable?

Steamhead

You're ahead of me there, Weez

I haven't tried them. We don't use zone valves much though, have found a dedicated circ on each zone works better especially on baseboard loops. I like to keep the flow constant thru the zone.

To Learn More About This Professional, Click Here to Visit Their Ad in "Find A Professional"