Buffer Tank Control

ALH_4

Yeah a 3/12 would take nearly 40 minutes to re-heat a 50-gallon water heater from 90°F to 140°F. That is probably not acceptable to most people, and buying two high efficiency appliances is probably cost prohibitive for most. Therefore manufacturers must have determined that the market for small mod/cons would be pretty small and possibly would generate complaints of inadequate DHW supply?

However, solar DHW pre-heat would probably help quite a bit in that situation. Then again, SDHW costs quite a bit more up front than a second fuel-burning appliance.

Then again, people seem satisfied with their gas-fired water heaters with similar recovery times to what a Vitodens 4/12 would provide, so why would they not be satisfied with a 4/12 and a 50 gallon tank?

I guess I'm off on a tangent, but it's hard to talk about buffer tanks without talking about the need for smaller boilers.

It's probably time for me to stop thinking. ;-)

Mike T., Swampeast MO

Do any of the available control systems "understand" buffers when used with and controlling the modulation of a mod con?

As I understand things, the only times you [might] need a buffer with a mod-con are when:

1) System load is significantly below minimum modulation.

2) The emitters cannot liberate the boiler's output at the minimum modulation level at the current reset target.

Number 2) has the potential to be the more troublesome...

Ideally it would seem, the buffer would be automatically charged whenever condition 1) or 2) occurs. It would then be automatically depleted as long as condition 1) or 2) continues. Otherwise, the buffer would "disappear".

I cannot see a "simple" way to do this as special hardware will be a requirement. I can though see it done with reasonably inexpensive and highly reliable components as long as the control exists.

Josh_10

I have taken a Tekmar differential temperature control and activated the buffer tank when the delta T drops below 10 degrees between the boiler supply and return.

Brad White

I have used a system return water aquastat

when the return water rises above 120F adjustable, the aquastat starts a circulator to charge the tank (which is otherwise off-line)

Brad White

I have used a system return water aquastat

when the return water rises above 120F adjustable, the aquastat starts a circulator to charge the tank (which is otherwise off-line). The circulator is on the first of a pair of closely spaced tees, the return from the tank goes in the other branch back to the boiler. I think of it as a thermal pole-vault.

The cooler buffer tank water continues the condensing as long as it can. Once the tank warms up the circulator continues to run to deplete it, using it instead of firing the boiler. I had a second aquastat to shut down the circulator when the tank cooled but it stopped working for some reason so I decommissioned that part. No time to deal with it I suppose.

When the HWR temperature cools, the circulator stops and bypasses the buffer tank.

A better way instead of the circulator, had I not put this together in haste, would be to install a motorized valve between two tees. When the HWR temperature rises, the valve closes, diverting all flow through the tank.

Still, a low loss header is to me the best way to get the coolest water back to the boiler. The buffer tank would then go in the return between the LLH and the boiler. But you knew that

I wish I thought of Andrew's idea though.

Paul Rohrs_7

Question Brad...

I am assuming that we are talking about a mod/con boiler setup with this buffer tank? If so, does the lower return water temp from the buffer tank present a "false" need for an increased boiler firing rate?

Paul

hr

Depends on how you pipe it

of course. As a secondary off a PS or separator the control would be different then it would in line on the return.

I like Brads idea with the motorized valve.

Although after watching that 80 gallon mod con HTP Phoenix work, in my shop...

Just buy a buffer tank with a modulating burner in it;)

I suspect reset control is or could be available for this unit.

hot rod

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

scott markle_2

load matching

Mike, it seems to me that your conditions(1and2) are different ways of saying the same thing.

As you know a house with carefully sized radiators can be heated very well on outdoor reset alone, no trv's, no differential bypass.

The degree to which an emitter can transfer energy is ruled by the same delta T principles that govern all heat transfer. Heat seeks equilibrium. The rate of Energy transfer is determined by surface area, convective influences and delta T. Between room temp. and water temp.

Seems to me that the the conditions you state are one and the same.A structures heating load is in a state of flux. The building envelope it self is a sort of radiator, again delta t, this time between inside and out.

An outdoor reset curve is about load matching, If our actual loads are lower than minimum boiler output than on/off operation will be required. Outdoor reset governs the amount of energy emitters can release. No mater what the surface area, an emitters out put is ultimately determined by this delta t relationship, or reset ratio in the case of a system so equipped.

I imagine this question may relate to the vitodens "pulse mode operation" you have observed in your "euro cave" experiments. My own house which is small and super insulated, experiences the same short cycle issues under low load conditions with a vito.

I am looking for a solution to this as well but as was revealed in a past thread this issue may have a lot to do with the burner operation. The N.A vito lights on high fire and throttles back to desired output. As a result, under low load conditions desired supply temperatures are often reached before the boiler even modulates down.

I don't think buffer tanks would be very helpful in this situation. Return temps. are not the issue. Under low load the burner shuts off before return temps even rise significantly. Even if a buffer tank worked it would blow the whole the compact low mass elegance of the system as far as I'm concerned. (My boiler hangs above a front load washer dryer on the second floor). I don't understand why these burns could not be longer and more spread out without fooling the system by means of a buffer tank. This is a boiler control problem if you ask me.

Brad White

You hit it, Paul- a good distinction to make

The buffer tank was installed on a non-modulating condensing boiler ("Non-Mod-Con"?), the Monitor MZ. The cycles were getting to me so I installed the buffer tank to slow down the bang-bang control.

You have an excellent point if the boiler uses return water as the sentinel point for increasing firing rate which would beget an elevated HWR temperature eventually.

The issue of cycling in low loads would be solved though! It would cycle of high limit! Interesting point you bring up, Paul.

Personally, I see buffer tanks as a fix rather than a positive design feature going in. It always seems to be the fix for low-mass boilers by defeating their energy benefit, that being low mass.

Uni R_2

Define control system...

I plan on using a monoflo on the return to the boiler with a 3/4" branch drawing from the bottom of a once electric water heater which is fed from the top from a very open 1.5" section through the P/S area.

With just under 4gpm flow that 40 gallon tank should perfectly match the 10 minute firing cycles on a Prestige. There'll be a valve on the outlet pipe to control flow.

The tank's top is designed to thermo siphon into the main heating loop in that P/S intersection area. It's piped across the top of the tank. The return's monoflo tee is near the boiler and it draws from the bottom of the tank - that is if it's clean enough to use. I think it's 5 years old but new one's are cheap.

The existing radiation here was supplied using monoflos so the ΔT is small - 6 or so. Matching the flow from the buffer tank at cooler temps over the 10 minute cycle will be an interesting comparison. Given the low ΔT of the system itself I think there's a good chance of getting stratification in the tank to help keep the spread wider than only 6°.

It would be nice to know the benefit and the true materiality of it. How many degrees before it's worthwhile? Is 2 significant? 5?

Uni R_2

Can I paraphrase?

"Personally, I see buffer tanks as a fix rather than a positive design feature going in. It always seems to be the fix for low-mass boilers by defeating their energy benefit, that being low mass."

I think Brad's trying to say that residential boiler's don't modulate down nearly low enough. That may because it isn't small enough or doesn't modulate enough.

Brad White

Pretty much, Uni

The smallest boilers tend to be too large or with few choice sizes for todays homes and yesterdays homes made tighter. Take away modulation, especially with low mass and hand me a tank.

ALH_4

Question

Am I missing something? Why control the buffer tank at all? Is there a reason to take it out of the loop? There are times when it might not be needed, but when would a well-insulated buffer tank hurt efficiency?

I wonder how much a buffer tank helps the efficiency and/or longevity with mod/con boilers.

Paul Rohrs_4

My take......

Is that I only add a buffer tank with the desired affect being to minimize short cycling. Extended run times in and of themselves will add to the efficiency when compared with a short-cycling system.

Generally speaking, I usually pipe them in series on the supply outlet of my mod/con system. Having said that, the very last system we did, we piped it as a parallel secondary loop that actuates any time there is a space heating call.

Regards,

PR

Brad White

My Take 2

I am a fan of keeping a buffer tank off-line when not needed. Paul's take about it making a boiler seek higher fire reinforces that for a reason I had not considered. Sometimes you do want lower mass for a fast response such as when you are picking up the load during morning warm-up. After that burst is satisfied the load is essentially done and it is time to coast -or cycle. By taking it off-line you can simply bring it in as needed and for "reasons specific" such as elevated return water temperature.

I prefer the return side to the supply side because, if it is on the supply side (AND in line without an option), you have to heat all of that water and run all of that colder water through the system before you get usable heat at the radiator.

By putting the tank on the return, whether off-line or on-line, you get "boiler fresh" hot water to your radiators right away.

By putting it in parallel as Paul just suggested, you can have the best of both but lose the ability to compensate for elevated return water temperatures. But Paul's suggestion does eliminate the "return water drop = firing rate increase" response at least.

Does that make sense?

Best,

Brad

Paul Rohrs_4

Follow up....

Totally agreeing with Brad here, but I would also add that
Brad's comment, "but lose the ability to compensate for elevated return water temperatures" might be arguable in that it would be at THIS time that the boiler modulation is downfired to a lower rate based on a function of Delta T.

Regards,

PR

ALH_4

Low Load = No Buffer?

But aren't the times when there are micro loads or an overall low heat load specifically when the buffer tank does the most good?

The more I think about this, the less I like the idea of a buffer tank with mod/cons. It's looking more and more like "stuff" that adds cost and complication without a clear benefit.

In my opinion, the only real solution to this problem is smaller modulating boilers to match small design loads or multiple modulating boilers to provide high turndown ratios for large design loads. We are all waiting for the former to be available. A 40MBH Prestige or a Vitodens 3/12 would be perfect.

I would rather let a Vitodens 6/24 cycle on a small load without a buffer tank than add a buffer.

Has anyone seen a case where the addition of a buffer tank to a mod/con has improved the fuel economy or function of a heating system?

Am I way off in my thinking?

Brad White

I see your point, Paul

It is a fine line but I see your point. We agree and are limited -nay thwarted!- once again by limitations in modulation downward limit. That return water cuts both ways does it not?

Paul Rohrs_7

Agreed



Pablo

Brad White

I do not think that you are way off, Andrew.

Yes, micro loads off of a buffer tank are an ideal strategy especially when you can positively direct your flow by small circulators or control valves. With an "open plan" and maybe TRV's (some flow to each but more or less, not "all on or off") then not so much benefit.

Agreed on a 3/12 Vito! Bungalow Heaven.

As for documented savings? That would go to cycle inefficiency. I for one lived for a year and a half without a buffer tank. I did not look for fuel savings specifically in that regard, I just wanted to stop the minute-on, 20 or 30 seconds off cycling, which it did. Saved where and tear on the motor and igniter I suppose. By this time it is hard to predict looking back if any savings. Too many other variables have since changed. Good point though.

Brad

ALH_4

hmm

I'm not sure what to think anymore. I have gone back and forth on buffer tanks. At one time I thought something like the HTP Phoenix would be perfect. Now I am not so sure. The low mass of most mod/cons is advantageous in thermally coupling the boiler to the system, but then there is the tendency to short cycle. I guess the question becomes, what is the effect of this cycling?

I cannot really think of a major disadvantage to having a buffer tank in series on the return on a low temperature system with outdoor reset other than the space used by the buffer tank and the need to purchase and install the buffer tank, associated piping and controls, if used.

I would be very interested to hear Viessmann's ideas on buffer tanks. There must be a reason they (or anyone else) do not offer a 3/12. HTP is close.

A couple of the many things I do not know about mod/cons:

How much fuel is wasted by the various mod/cons during ignition?

How many cycles are the components designed for?

hr

I've done 2 MZ's with buffers

right on the return to the boiler in series with the radiant manifol. The only time it did much "cycle extending" was on the first cold start up. Once the buffer got up to the boiler temperature it went back to short cycling. I tried one with a 30F delta T setpoint control as the boiler start. Better for a big shop but too wide for comfortable residential.

You really need to be able to couple and un-couple the buffer, in my opinion, with some sort od algorithm control as Mike alludes to. IF you want ultimate performance

hot rod

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

Brad White

You raise good points, Andrew.

The only reason I think that manufacturers do not offer smaller mod/cons is the domestic HW factor. A small boiler sized for a 30k heat loss has little left over for DHW production.

If it was coupled to an indirect you would take a much longer time to heat the water and the space would suffer. The solution would be a separate stage for DHW but then why not keep a separate DHW heater? I would settle for a "heating only" mod/con at 35-40 MBH input with separate DHW.

Your ambivalence is the reason why I favor decoupling the buffer, take it off-line when you do not need it and kick it in when you do. Naturally this means longer run-times but you also get the benefit at the other end while the boiler rests, you draw off your buffer. It is not free so much as deferred and your boiler gets a break.

As for cycling waste, with low mass naturally there is less time spent "off" (or "on" for that matter) but you still have that period of ramp-up time that has you finding stable combustion. Most of your CO is produced in the first two minutes; while not a strong issue it is a sentinel of incomplete combustion.

Component wear is my stronger concern. Everything has a lifespan. I am with you.

scott markle_2

control

I don't understand why a sophisticated boiler control can't permit wider differential under low load to prevent short cycling. Tekmar controls incorporate this feature. And from my limited observation it works well.

The vitodens is unique (as far as I know) in that it is able to maintain supply temperatures accurately enough to maintain even room temperatures without any indoor feedback, and without mixing or injection. Perhaps the short burns under low load are necessary to accurately match load and output. Especially given the throttle down operation of this boilers burner. Personally I would rather have some loss of accuracy under these conditions than pay the mechanical costs of short cycling. No mater how well made the components I can't believe that this type of operation will not have a impact on longevity.

But to fool this compact elegant boiler with a bulky buffer tank does not seem like the solution,and I'm not convinced it would work no mater how intelligently it was implemented.

The point made about minimum boiler sizing being based on DHW requirements is interesting. Ideally solar assistance could make up the difference, as opposed to the redundancy of a separate water heating device.

I believe a european market vitodens with built-in storage tank and solar coil is available that fits this description.

Brad White

Fair enough

I did install the MZ version of their OD reset control, an entry level Tekmar unit and it did just that, widened the differential. The buffer tank made a difference, the control improved it by about another unscientific 30% or so. Just seemed to cycle even less.

I think the two go hand in hand, especially if the buffer tank can be switched on and off line.

Great point, Scott- it did fill in the entirety of my personal experience on this subject. Thanks

Unknown

I see two scenarios only.

1. You are cycling. Buffer tank extends your cycles. All water should be heated maximally to maximize buffer capacity, if you are in a system that should use a buffer.

2. You are not cycling, with your modulator.. buffer tank is invisible to the system. no reason to take it out.

I don't see any reason to use a buffer tank as anything except a hydraulic separator on a mod/con system. in a conventional system, where you need boiler protection, more is needed. But otherwise, I don't see the return advantage. What if your "boiler fresh" water stops the heat demand too fast to heat up the buffer tank? Then you defeated the entire purpose of having it there in the first place.

ALF

I've been struggling with buffer control for months. In the geothermal market, we cannot modulate output, we have a low mass single capacity heat source. In low load conditions, the buffer tank provides run time. What I want to do next, is draw from the buffer tank to provide heat during the off time of the source. I think that the buffer tank feed can be controlled as a secondary loop using a Taco pump with a differential control built in using the delta T between supply and return. I am presently building a test bench to prove if this is applicable. Has anyone else been down this road?

Unknown

why not just use a boiler control that can do reset, give you a firing differential, and decouple heat demands from unit runtime?

Assuming that the radiant or heat emitters can use mixing, you can even do this with the mixing control, perhaps.

This way it will only heat the buffer tank incidentally, when the system really needs more heat. The buffer does what it is supposed to, and just provides passive mobile mass storage, without being a STORAGE tank which is maintained at some arbitrary or too-high of a temperature.

that's just my thought.

I have two questions for you though Al:

1. Is run time itself really important for geo units, or is primarily just the "no heat" interim period that is a problem (I know some of them won't fire for ten minutes after they are done firing)? Cause if it's just cruising though the delay period that is a concern, you just need size a buffer tank based on the demand over that period of time, corrected for the demand itself increasing your run time while it reduces your buffer capacity's time to drain down, and the temperature differential you want to run from start to stop...

2. If run time is important for EFFICIENCY... can you explain why in simple terms? kw draw doesn't change.. there is a short period of time before the "heated" or "cooled" fluid makes it back from the wells or loop field... is that is? Or is there something else involved?

we scratch our heads at geo too a little. you guys and your magic ground heat

ALF

Like any device, there is inefficiency in short cycling a machine, the source loop reall does not effect the decision. what we are getting into is that people now want to heat and cool with the same unit using hydronics. The old drawings for geothermal show the unit pumping to the buffer tank then the load was attached to the buffer tank. This meant to go from heat or cool you had to condition the entire tank. Lots of lost BTUs. I have designed a piping diagram using primary secondary loops. The tank is a secondary loop. The primary loop will be controled by an outdoor reset device, that means during most of the year I will not be at design temp. I can run the buffer tank loop at a steady low flow rate, that means even at max load I have to condition the buffer tank. I want to shut the buffer tank loop off entirely if the load matches the output. This is rather primative, but better than what we do now. Ideally, I want to very the flow to the buffer tank to minimize adding load under full load conditions. Its purpose as a storage vessel is less important to me than its purpose to prevent short cycling. Make any sense?

Unknown

conditioning the buffer tank at max load is "invisible". It's not like the tank is cold and you have to heat it up.. and then lose the heat. It's just more water mass and volume in the system and all it does is charge up with heat and draw down AS the heat is used. If you were perfectly sized, you'd still just be replacing the heat you use.

I don't see any reason to ever take it out of the loop unless you're trying to bounce back and forth between heating and cooling, and if that's what you're doing, I'd say you're going to have efficiency issues whether the tank is there or not, because even without the buffer you still have an entire hydronic system full of water... never mind any further mass in the system like concrete, etc.

Unknown

geo buffer

ALF: the geo buffer tank strategy we've employed successfully for 2 years now is thus: buffer tank piped much like a low-loss header (primary pump from heat pump into tank & secondary pump from tank to system) with a tekmar ODR control (256,253 etc.) set up with a constant demand (jumper T to C) monitoring the tank temp. thermostats are just "taking" heat or cool water from the tank as they need it. this method gives us nice long run times and prevents short cycling via thermostat zone overlaps. also i've found that you can get away with using much smaller buffer tanks with tekmar controls because of the automatic adjustable differential they employ.

ALF

What we are beginning to see are larger loads requiring more than one unit. This makes the tank much more expensive to get one with large enough ports to handle sometimes 45 gpm and above. Also, you can't get to cooling with this configuration without conditioning the tank. I agree, for one unit heating only this arrangement is simple and it works.

Paul Rohrs_4

Tekmar TN4

I would really like to see a TN4 arrangement for all geo equipment to include contacts for fan-coils, 230V(ac) pump-paks, zone circs, and contacts for the reversing valve.

Regards,

PR

Unknown

geo tn4

i've been told by tekmar that it's coming. thermostats with radiant cooling functions are also in the works too i've heard.

i just attended one of their courses a couple months ago and went on a quick tour of their mech room, they had all kinds of one off controls: "tekmar heat pump control" & "tekmar ground loop control" among them.

jalcoplumb_7

Buffer w/ injector

I have used this layout with great success. I use a Tekmar 361 injection control. I put the boiler sensor in the buffer tank. The Munchkin has Vision 1 installed to handle the domestic load.

When the injection control fires the boiler, it activates the Munchkin. I reset the boiler on a higher curve than the system. I use this when I have a very small loads.

The boiler modulates and has a great run time. The run time is helped by the reset curve

Josh_10

Thats good to hear. Hopefully they come out with em soon as we need the controls NOW.

We have been using HBX for ground source. With good luck I might add.

Hopefully Tekmar doesn't promise us a control and then takes 2 extra years to release it like the TN4 internet gateway. Which by the way I am red hot about after promising it to my customers by spring 2007

Unknown

I'm ticked about that too Josh, but that's a good lesson now that controls are becoming more like software... it's "vaporware" until it's heading for the shelves, and the only release date is "when it's done".

Paul Rohrs_4

The only bad thing about the Tekmar Gateway

was that I was in TN4 training 1½ years ago when it was just about to be released and I was given an "overview" on it.

Maybe something to see at REX next month?

Regards,

PR

Rob Blair

Just attended the tN4 class in Cleveland

You are looking at 2008 for the Gateway.

rob