ECM and Converted Gravity System

SWEI

Gordy said:

Direct pipe not good no matter the flow. for a gravity conversion.

Works OK on a fire-tube mod/con IME.

Steamhead

The main differences between pumped flow and gravity flow are, as stated earlier, the tendency for gravity flow to heat the top-floor rads first, and the fact that gravity flow will change with how hot the boiler is.

The former characteristic resulted in several different ways to equalize the flow between upper and lower floors: using smaller pipe sizes for the upper floor rads, taking lower floor rad runouts off the top of the main but taking upper floor ones from the side, using orifices on the upper floors etc.

The latter one you really couldn't do much about. If you weren't circulating enough water to reach those last few rads, they just didn't get hot. Your only option was to run the boiler at a hotter temperature, boosting the flow rate.

What we're trying to do on these gravity conversions is maintain the same flow rate you'd get with the original boiler at its hottest, regardless of the water temperature. If you go too low, those last rads won't heat, no matter what your ΔT is. So we typically use properly-sized single-speed circs on these jobs, which locks in the flow rate, and deal with the water temperature by using ODR or some sort of slope control like a HydroStat.

ECM, ΔT, ΔP and similar circs certainly have their place. But I question whether they are needed on a gravity conversion job where the flow rate is pretty much constant- unless you like to experiment with them. Then, by all means go for it.

Steamhead

Paul48 said:

Steamhead......Would re-piping this reverse return be of any benefit?

Probably not enough to justify the cost. Once the circ is properly sized, the system should heat evenly.

Gordy

null

Was referring to ci boiler kurt

SWEI

Gordy said:

Was referring to ci boiler kurt

Check, understood. I know you meant that, but...

Gordy

null

I know I know.

Steamhead

Hatterasguy said:

Looking at your chart for a system with 640 EDR, 96MBH boiler, one might expect a flow rate on the original boiler of something around 5 GPM on gravity feed. Could even be less. The flow rate probably cannot keep up with the boiler (the boiler never runs continuously).

The recommendation in the chart for this boiler with a high mass system is 15 GPM..............way above the flow rate for the gravity system at its hottest.

So, clearly, you get into problems with pumped flow rates between these two extremes. One way out is to overpump to develop resistance in the earliest risers.

However, this can get costly if a big pump is required.

Could you not simply utilize valves on the risers, or the radiator valves themselves to achieve the same end without the need to overpump?

That is simply a function of what circs were available at the time I did the chart. It was intended as a quick reference that would let you make a selection without grossly oversizing the circ.

The basic concept comes from the old Bell & Gossett manuals, published when many gravity systems were being converted to forced circulation. They established flow rates and circulator sizes based on the amount of installed radiation, then added some because of all the water in those big gravity-sized pipes. When I experimented on my own system (that's right, Steamhead has a converted gravity system in his own house) and later some other systems, I found that if you got much below those flow rates, the ends of the system wouldn't heat well. So it was better to go a bit over than under, as everything would heat reliably.

Don't get too hung up on flow rates or ΔT on this type of system. You're not changing the system piping, so these are going to be what they're going to be. A flow rate slightly above the theoretical optimum will work fine, certainly better than if it's two or three times that optimum rate- which I find more often than one might think.

hot_rod

Maybe consider a B&G Vario ECM. No bells and whistles, just an easy to adjust ECM circ. I think the dial goes from 1-10, plenty of range.

Steamhead

Does it go to 11?

;-)

Harvey Ramer

Hatterasguy said:

Steamhead said:

Don't get too hung up on flow rates or ΔT on this type of system. You're not changing the system piping, so these are going to be what they're going to be. A flow rate slightly above the theoretical optimum will work fine, certainly better than if it's two or three times that optimum rate- which I find more often than one might think.

The theoretical optimum flow rate would be exactly what the boiler can deliver. In the case of the 96 MBH boiler, you'd be looking at a bit under 8 GPM for a 25DT. Per the chart, you've now doubled that to 15 GPM to get it to work. I see this as 2X the optimum rate.

Would it not be preferential to put some restriction on the risers or use the radiator valves themselves to control flow in the first few radiators if the last one is problematical?

I understand that the big pump can be used to develop sufficient restriction in the risers of the first few rads, but is this the optimum approach?

I would like to interject something here. The optimal flow rate on an emitter system is whatever keeps the people comfortable for the least amount of money. That includes both installation and running costs.

The optimal flow rate on a boiler really varies dependent on the boiler construction and design.

The optimal flow rate for a boiler and the emitter system are not always the same.

icesailor

Say, did they put orifice's into gravity coal systems? I never saw any. Maybe someone slipped up. I've seen them in later pumped systems. They were a giant PITA to slide them into the unions and not have some of them leak. And they still worked like a dog that just rolled in a dead animal.

So why does this work (or in the case of some, why does it not work.

You take a converted coal to oil gravity system that is now pumped. The boiler runs up to whatever high limit it can get to before shutting off. Its cold start. Its got all GBP's, everywhere and GBR's too. And not an orifice plate in the entire house. Some oil Peerless oil boiler fired at 1.75 GPH. The upstairs back radiators in a bedroom never get hot, barely warm. The kitchen under it barely gets warm. The first radiator above the boiler gets so hot that its hard to keep paint on it. It satisfies the thermostat and shuts off the pump.

Install the 4-way mixer. The 170+ degree water circulates on one side of the valve, the ODR and low temperature water circulates on the other. The thermostat tells the burner and the primary pump to start. The secondary pump starts whenever the primary pump runs. It senses the return water temperature and opens that 4-way to allow hot water in, controlled by the ODR sensor. The valve also senses the water going in to the boiler so it doesn't get cold shocked. The temperature slowly rises in the whole secondary side. If it needs 125 degree water, for some reason, after all that water runs through all those GBP's, and all those GBR's, it only has a return temperature of 110 degrees. But in a few hours of starting from cold, all the radiator inlets are around 125 degrees. Valhalla!! And that is on the second speed of a Wilo 3 speed. Oh yeah. That overheating front hall radiator had 125 degree water going in it.

Works just like the 1910 Gut Rehab that was done when they put in the system from scratch.

Sounds like Siggy's "Equilibrium.

FranklinD

As my boiler sits in it's present state, it is direct-piped. Thus my futzing around with a variable speed circ and too much time spent considering the optimal flow rate.

For the boiler, it should be about 7 gpm max for a 20* temp rise.

For my emitters (at a design temp of -20*), the flow rate for a 20* temp drop should be 7 gpm (total system) or so.

Perfect match? Sure, except that to heat that last radiator, I have to raise the flow by 50%. I would love to run the circ at 7 gpm, in fact I did for the first 2 months of the season. Only issue was that my family lived in the second floor bedrooms because it was cold downstairs.

I agree that I should probably be able to run the system satisfactorily at a low flow rate, but at this point, the living room is warm, and that is the critical factor to my wife and kids. And the Bee only consumes about 28 watts running on the speed it's at. So I'm okay with it.

I've picked up even more info from this thread and want to express my appreciation to the pros involved. You folks are something special.

Abracadabra

Steamhead said:

Does it go to 11?

;-)

LOL! I'm surprised no one picked up on that

https://www.youtube.com/watch?v=KOO5S4vxi0o

FranklinD

Hey, I laughed out loud a bit when I read it.

Steve Thompson (Taco)

Gents, this is a great post... This is my takeaway (for sure, correct me if I'm wrong - not a gravity expert)

The boiler is 135,000 BTU's output - at 20 deg delta T (this is what I might be missing) the system needs about 10 USGPM @ 3.5' (read in a previous post). The low head makes sense with a gravity conversion (low water velocity). The most important part of this is remembering this system has an extremely flat system curve so if you want to use an ECM on constant speed it needs to have a flat curve. Better yet set the ECM circ to constant pressure (high system friction loss needs steep curves). It wouldn't take too much "excessive circ head" to over pump the system.

I can't be 100% for sure but I believe a delta T would work well in this this application as it's "curve" would be what the system needs (same BTU train thing). I doubt if a delta P circ would work as it would see very little, if any pressure changes as zones open/close/modulate doe to the flat system curve.

Question: Do we know what the design delta T was back when this was laid out? Or was it "what it needed to be"?

Lynnwill

Steve,

You are correct. This post has been very informative.

In answer to your questions: The house and heating system was constructed in approximately 1930. There is no way to know what the very dead men designed. Given that there was a 330MBH boiler when we arrived 17 years ago I assume it was meant to just overwhelm the heat loss when fuel was cheap.

The existing 9 yr old boiler is indeed 136,000 DOE output.

BTW..I have settled on your VT2218 ECM. It is not widely available but I found a source. Looks like a great product and an improvement over the Bumblebee.

icesailor

@Steve Thompson (Taco):

I'm quite certain that if you asked any of the really old dead guys about "Delta", they would have told you that it was the place on the Mississippi River Delta that the Acadians moved to when the British kicked them out of Canada. Or, just a place in Louisiana where they play good music.

Gordy

I dont know if there is a definitive answer other than what was used to design it.

icesailor said:

@Steve Thompson (Taco):

I'm quite certain that if you asked any of the really old dead guys about "Delta", they would have told you that it was the place on the Mississippi River Delta that the Acadians moved to when the British kicked them out of Canada. Or, just a place in Louisiana where they play good music.

Maybe in their math there was a delta. But that very delta was variable just like the flow. They did what technology try's without electricity. Mod con coal boiler with variable output.

vicbrick

Greatly appreciate the information/comments shared in this thread. I am also struggling with optimal flow rates/pumping for a converted single zone, 2-pipe, gravity hot water system:

11 CI Rads with ~830 sqft of EDR (2 1/2" supply/return mains)
Calculated heat loss of 140,000 BTU/hr at 2F

Would like to replace current 35yr old, 240,000 but/hr boiler/B&G series 100 pump with a 150,000 btu/hr Dunkirk Helix VLT (P/S pumped) and a single Grundfos Alpha 15-55 pump. Will add 3 -4 TRV's to the radiators during the boiler upgrade...hence, the desire to use the Alpha.

dicey part:
Would like connect the new mod/con boiler to the existing 2 1/2" ret/sup mains via a 1 1/4" secondary circuit (~12ft) which will contain the Alpha pump, DirtMag,etc. The 110 yr cast iron mains have existing/unused 1 1/4" ports on both the return and supply...would love to use these, rather than break into and replumb the old cast iron. So, I would have a pumped primary circuit contained inside the Dunkirk VLT, and the Alpha pumped secondary circuit which connects to the supply and return of the existing single zone.

Doable?

Any thoughts or concerns, etc. would be greatly appreciated.

Thanks!

Steve Thompson (Taco)

Curious - Why use a Delta P circ and not Delta T? Taco has both so I'm not biased here - just askin...

Gordy

I thought I read somewhere that a delta t circ will actually act like a delta p. Flow will slow as zones close trying to keep that delta no steve

Steve Thompson (Taco)

They "kind of" act the same. Loads decrease they slow down. But delta T reacts to changes in delta T. Delta P reacts to changes to the torque on the impeller (push back as zones close slows the circ down).

So if this application will have the delta P "circ" see the pressure push back then it will modulate and help get the delta T close.

This is why delta P will not change it's speed when used as a zoning circ. No change in head "push back".

Gordy

Not really implying one takes the place of the other. I think in a zone valved system to use a delta t in place of a delta p needed pump the delta t would be slower to react than a delta p pump, but it would work.

Now in a gravity Trv'd system hmmmm. How much "push back" does the delta p circ need to react. How much will there be in a gravity system with little head to begin with. Depends how it's piped. Trvs would have bypasses so next rad in series would be able to get flow. So the end rad in the circuit would be the one doing any restricting when it's Trv modulates closed. Key ones being the last in the series.

Then the fact that most gravity type systems , and the associated dwelling they service. It is unlikely that a large amount of trvs would be closed partially, or fully at the same instant due to the older structures heat loss.

So in reference to the last new posters gravity system answer I would says delta t would be my choice. Really it will cover delta p in that scenerio. A little slower to react, but as trvs close delta t would narrow, and there for pump speed would slow to maintain the delta t setting.

Be nice to have a circ with built in delta p, and delta t. If for nothing else just to,say you did it.

Harvey Ramer

I have stood at a pumped gravity system with my hands on the pipe as it slowly warmed up. One thing I noticed was that there were considerable fluctuations in temp. In other words, I'd feel eddies of warmer water flow through the pipe and then a bit cooler and then warmer and then cooler... kind of random.

Not saying this is so on every system, but on this particular one a delta-t would not do well.

If only putting trv's on a couple rads, there is no need for a delta-p or delta-t pump. Nothing really to be gained.

Fixed speed is usually best for gravity, from what I have seen.

Gordy

Thats what I was visioning in larger pipes a swirling mix of hot, and cold water.

Paul48

That has to be frustrating for the OP. These last few posts seem to ignore every problem discussed in this lengthy thread. I know I said WTH to myself.

Gordy

I think there was enough information to the OP to make an informed decision. I dont think breaking off on a tangent is a terrible thing if the discussion lends itself to some insightful theories, and observations. If the OP is confused then he should come back.

icesailor

Harvey Ramer said:

I have stood at a pumped gravity system with my hands on the pipe as it slowly warmed up. One thing I noticed was that there were considerable fluctuations in temp. In other words, I'd feel eddies of warmer water flow through the pipe and then a bit cooler and then warmer and then cooler... kind of random.

Not saying this is so on every system, but on this particular one a delta-t would not do well.

If only putting trv's on a couple rads, there is no need for a delta-p or delta-t pump. Nothing really to be gained.

Fixed speed is usually best for gravity, from what I have seen.

I've done that with many a pumped regular and gravity systems.

I noticed that wavering on pumped gravity systems when starting from cold. I also discovered that on systems with a 4-way and ODR, it doesn't happen. After a while, it all balances out. No fluctuations.

Paul48

Maybe I'm confused, but the information I thought I had gleened from this post, was, that minus a way to properly balance flow to each radiator in a converted gravity system, ECM circs were not a good choice.

Paul48

Which would lead to sequentially heating the structure, as opposed to evenly heating the structure. If that is not a problem, then go for it.

SWEI

Gordy said:

Be nice to have a circ with built in delta p, and delta t. If for nothing else just to,say you did it.

The MAGNA3 does this now. Again, can we please have some smaller models?

Paul48

Hat.......You mean like it was, when momma was a stay at home mom, and grandma and grandpa, lived with them? Dad or the kids stoked the boiler, before work or school. I think the expectations have changed enough that it may not be acceptable by todays standards.

vicbrick

Sorry, if I hijacked the thread. Have been reading the forum for a while and this thread had attracted a lot of valuable input from the "seasoned veterans". greatly appreciate the feedback/input!

It appears that two working solutions to gravity conversions exisit: manually balance low flow conversions with piping/valves or "overpumping" .

My existing system is overpumped (27ish gpm), but the radiators all heat uniformly well and system is fairly quite. Hence, I would prefer stick with "overpumped" solution to move the heat around and try to sort out improvements/upgrades to the boiler and pumping method.

To me the additional cost of the proven Alpha pump is more than offset by the additional flexibility to adjust it's operation in the installation. So, I if want to try optimize the system performance at later date, I will be able to easily.

just not sure the piping and alpha pump arrangement I outlined in my earlier post will allow me to achieve flow rates close enough to the current overpumped system to perform well. Think I might be end up a little on the low flow side...high teens? ...with 20 gpm as a target

and yes, the three TRV's I am planning on installing are on the end of single zone circuit.

any thoughts or concerns are appreciated

Thanks

SWEI

Proportional zone valves don't have issues with this.

Paul48

Wouldn't it be easier to deal with 1 fitting and riser. Hang a 2" pipe, and make it reverse return? I think that would eliminate having to overpump it. If you're really lucky, the end of the return main might have a plugged tee.

Gordy

I think I would rather use a Trv. With constant circulation. The end result would be far better.

Paul48

I've had the pleasure of replacing some valves on my 90+ yr old radiators. I'll pass!

SWEI

Hatterasguy said:

Why would you use a proportional zone valve when a simple butterfly valve would serve the purpose?

I recently found a butterfly valve with an angular scale (0-90 degrees) so you can tell where you are and where you might want to go. Not too costly, either.

Take a bit of experiment to get them all set properly, but once you're done, it's perfect.

I was alluding to a proportionally actuated valve for variable flow control in a zone. Sounds like you're describing a manual balancing valve. Both have their places.