Pump Speed

[John Molyneux]

My system is piped primary/secondary. There's a 3-speed pump inside the boiler that I have on its lowest setting. I have a Grundfos Alpha that supplies water to the zones that I've been running in 'auto-adapt' mode.

When the system is stable at near-design outdoor temps the Alpha indicates it's pushing about 3 gal/minute with 13 Watt energy use to the primary zone (the second zone is much, much smaller). Under those conditions it appears the boiler pump is running at a slightly higher speed than the Alpha. In other words, I can detect some (but not much) 'backflow' across the primary/secondary header meaning that some of the hot water coming out of the boiler is being recycled back into the boiler and is therefore diluting (slightly increasing the temp of) the return water from the radiators before it enters the boiler.

Is it correct to assume that the flow through the pri/sec header should be neutral, or at least mildly positive toward the radiators, such that the boiler is seeing the lowest return water temps it can? Should I manually bump up the speed of the Alpha?

Thanks!

 

[Leon82]

In a perfect world, yes.

but speeding up the alpha pump could cause the return water temp to rise a little because less heat would transfer from the water to the radiators, I think lowering your delta t.

 

[Sponsor]

 

[BadgerBoilerMN]

Smart guy.

Richard Trethewey wrote a short article about primary/secondary pumping in regard to ModCon boilers complaining that the concept was primarily used to "keep boiler hot". I haven't heard anyone, but me, complain about it since. The reason is simple. Few care and fewer still will do the math--if they know how.

Kudos.

I have an Weil McClain Eco70 on a multi-zone radiant system driven by one (1) Alpha, 35 degree DT max.

Don't tell anyone...

What boiler do you have?

 

[John Molyneux]

Leon: That's a good point. And I'm so optimized at this point I'm definitely in the area of diminishing returns. Hit my design day last night and the boiler was humming along at low fire sending 108 degree water out to the CI radiators. It's been condensing like crazy all winter.

Hi Badger: I kinda figured you'd weigh in with the 'one-pump' thing. I don't think my Bosch ABC certified installer was willing or able to go there, though. Plus, I kinda wanted them to install it per the manual.

It's a Bosch Greenstar 57.

Just curious: what's more important -- some specific Delta T or just getting the lowest possible return water to the boiler?

 

[Dana]

The entering water temp of the boiler is the primary determinant of combustion efficiency. But if your output temp is 108F you'd already be in the high-90s for combustion efficiency even with a 2 F delta-T. Dropping the return water another 10F with a bigger delta-T may only buy you another 1% in combustion efficiency.

Boilers also have a minimum flow rate requirement necessary for avoiding flash-boil issues. Combustion efficiency will drop due to poor heat exchange if a layer of tiny bubbles is allowed to form between the slow-moving water and the heat exchanger. You can pump it a bit slower than spec if you raise the pressure of the system to suppress the boil.

Most mod-cons will tolerate delta-Ts up to 50F without a problem. (Tankless water heaters are designed to tolerate 100F+ delta-Ts.)

 

[John Molyneux]

Dana: can you help me confirm my year-over-year comparison now that I have a couple months of heating season data with my new boiler and some envelope upgrades?

Based on your suggestion, last year we used an 85% efficiency factor for my old boiler and Base 60 weather data since we'd been keeping the house on the cool side (approx. 63F). This produced 450 BTU/degree-hour and a total implied heat load of approx. 30,000 BTU/hr. (Interestingly I subsequently had an energy audit with a Manual J, which came in at 29,000.)

This year I used 189.6 therms of natgas between meter reads on 12/03/15 and 02/03/16. There were 2,005.8 HDD during the same period at KPWM.

I'm assuming I should use a 95% conversion factor with the new modcon and that I should use Base 65 weather data since we're keeping the house warmer this year (approx. 66F). Is that the right way to think about it?

If I do the math that way I get 354 BTU/degree-hour and a total heat loss of 22,324. (It's also interesting that my energy audit predicted we could get my heat loss down to about 23,000 if I implemented his recommendations, which I did.)

So, it looks like I have a 25% decrease in my heat loss. And I assume there's some extra energy savings that aren't included in that due to overall efficiency characteristics of the boiler?

 

[BadgerBoilerMN]

Bosch does not require P/S piping or pumping, though some manufacturers do. This boiler has a low-temperature output of 122F of <55mbtuh the required flow at 40F is <3gpm. With an integral pump with 4 speeds up to 7gpm at 10 feet of head I think you would have to have unusual distribution or radiation to require a second, parasitic, pump.

There are certainly diminishing returns on low water temperature operation but the best condensing boilers, in fact most, control combustion efficiency in a two-chamber approach supporting "hot" combustion in one while bring byproducts to dew-point in the second. Most will start making some noise--flashing to steam--at 40F or so and certainly not a place you want to be in terms of efficiency or more critically the erosion that can occur when heat transfer fluids boils and implodes on the waterside of a heat exchanger. Modulation has really killed this problem except where the ill-informed over-size boilers and insist on micro-zone fire control.

The practice of P/S pumping can only be defended in certain retrofit applications, where anti-freeze MUST be used and where people willing let others do their thinking for them.

It is undeniable that the additional pump requires that both operate to get heat and that the extra doubles the electrical consumption and the statistical probability that a pump related no-heat call will be in your future.

 

[John Molyneux]

I've read in the literature that Bosch doesn't require P/S for the Greenstar only if it's feeding a relatively small single zone baseboard system and that they require or at least prefer P/S in many situations. This is what my manual says ...

"Primary-secondary piping or a low loss header

Radiant floor, panel radiator, multi zone and application systems exceeding 70' (21 m) of 3/4" baseboard at a 20°F delta T and a flow rate of 4.0 gpm plus the necessary supply and return piping, require primary secondary piping or a low loss header for hydraulic separation of the system from the boiler."

I'm definitely not suggesting you're wrong about the virtues (or lack thereof) of P/S -- just pointing out what it says in the book.

 

[BadgerBoilerMN]

I just do the math. Not as fast as Dana or his brother Cray, but I get by.

 

[Dana]

Typically the heating/cooling balance point of 2x4 framed house comes in about 4-5F cooler than the interior temp. The 4-5F of delta is supported by the combination of plug loads, mammalian & avian occupants, etc. A single sleeping human puts out 225-250BTU/hr, which is getting onto 1% of your estimated load. Refrigerators and other 24/7 plug loads also add up to real percentages. Even the pumping & control power used on the heating system is likely to be covering a percent or two of YOUR load. So if you're keeping it at 66F indoors running the heat load calc numbers at base 61F or 62F, makes more sense than using 65F.

If your AWT is always under 110F even on design day (sounds like it is) you're probably averaging a bit higher than 95%, but measuring that with any precision could be tough.