Boiler readings don’t jive with the math

TommyA

I’m hoping for some expert opinions on where my discrepancy is. Here’s my situation… I have primary/secondary piping with the primary consisting of a Weil-Mclain CGa-3 installed last season, Caleffi 280 thermostatic mixing valve (130 degree cartridge) and a Taco 0018e variable circulator on the return pipe. The secondary side consists of big pipes and radiators (old gravity system) with the OEM Taco 007e.

When the boiler return water temp reaches 130 degrees, the Caleffi valve partially opens beginning to blend in the cold secondary water. It will maintain this return water temp for an extended period of time while the secondary water volume slowly heats. The boiler supply will hover around 148 during this time (a sort-of equilibrium).  This is when I’m taking my measurements across the boiler. Here are the readings:

• Gauge at boiler input/return: 131 degrees F.
• Boiler digital display: 151 degrees F. 
• Gauge at boiler output/supply: 148 degrees F.
• Flow / Head: 2.89 GPM / 2.95 ft. (According to the Taco 0018e bluetooth app)
• Boiler ratings: 67000 Btuh input / 56000 output / 48000 net / 84% eff
• Gas meter consumption: 57 seconds per cubic foot (63158 Btuh)
  

The readings do not jive with the equation: GPM = Btuh / (500 * deltaT). Is the boiler less than half as efficient as rated? Is the GPM reading completely fictitious? Are the temp gauges way off? All of the above? Something else?

Wondering what you all think is going on here. 

Thanks, Tom

hot_rod

the readings of flow on m many of the small ECM circs is a calculation, not an accurate flow measurement. I believe power draw and rpm are used. I have seen as much as an 80% discrepancy when measured with lab quality, certified flow measuring devices.

The type of fluid, and air in the fluid stream can throw off the numbers.

Some circulator models and brands have small flow sensors mounted in the pump body. They use a vortex shedding technology. Those too can be a little inaccurate.

If you can develop an accurate system curve, lay it over the pump curve, you can nail down an accurate operating point and get a better idea of the actual flow rate to use in that equation

A module in the HDS software can develop the OP, and an find example of a long hand calculation in Idronics 16

IMG_0976.png

IMG_0977.jpeg

mattmia2

If the gas meter is accurate(which is a big if for a lot of reasons) it is also about 10% underfired which won't help with heating the large mass of water. Might not be a bad idea to get someone good with combustion to take a look at what is actually happening.

jesmed1

Also, that 84% AFUE is a dry gas efficiency that does NOT account for the additional roughly 15% latent heat of vapor loss in natural gas combustion. So your actual combustion efficiency, including latent heat loss, is more like 70%, giving you an actual output of around 47,000 BTU/hr, which you can then apply your 15% piping loss factor to if you want to estimate how much heat is actually reaching your living space.

But because you're measuring conditions right at the boiler output and return, the 15% piping loss doesn't come into play for this calculation. But as it turns out, the 15% piping loss you included to get your 48,000 BTU/hr "net" equals the 15% latent heat of vapor loss that you didn't include. So your 48,000 BTU/hr net turns out to be the right number to use, but for the wrong reason.

Jamie Hall

Ah… no. Boiler — or combustion — efficiency measurements are, I'll grant, difficult to make with any accuracy. However, the loss from the latent heat in the flue gas is actually the reason that the 85% figure can't be improved without condensing In fact, recovering that lost heat is how a condensing boiler can achieve higher efficiencies.

To say that one must then double subtract that latent heat from the nominal efficiency is misleading — at best. It's already been subtracted once.

Now this is not to say that the real world overall efficiency of a boiler is what is measured with a reliable combustion test is also somewhat misleading — just as the real world fuel efficiency of a vehicle is not always what's on the window sticker. "Your mileage may vary". The overall efficiency could be greater (unlikely, but theoretically possible) or worse (sadly, considerably more likely), but that is overall system efficiency, not combustion efficiency.

EdTheHeaterMan

Or the math could just be wrong. You know how they teach that new math these days!

mattmia2

https://forum.heatinghelp.com/discussion/comment/1836496#Comment_1836496

it was the way i figured out to do arithmetic on my own because the early 20th century algorithms that they tried to teach though simpler to explain were much harder and more cumbersome to actually calculate.

Jamie Hall

And another little twist. It seems to be rarely appreciated that the latent heat which is not recovered in a non-condensing application varies with fuel type. That potentially recoverable heat is higher with natural gas than it is with LP gas (butane/propane mix) — and very much higher than is present with fuel oil (which can be approximated as decane). Pure hydrogen has the highest difference (around 18%). Natural gas is around 14%, and fuel oil is around 6%. (those percentages assume that one is able to utilize the latent heat released on cooling to 25 Celsius — 77 F — which is simply not feasible in a space heating application — cooling to to 120 or so will return about half that).

jesmed1

https://forum.heatinghelp.com/discussion/comment/1836494#Comment_1836494

Jamie, I respect your credentials as a P.E. Unfortunately, boiler combustion efficiency numbers have caused a lot of confusion on this forum and in the profession.

I did my Masters work in mechanical engineering on the combustion of rocket propellants, and I've investigated how boiler combustion efficiency measurements are taken and reported. And all combustion efficiency measurement devices used by the boiler techs measure only "dry gas" losses. They do not measure latent heat of vaporization losses. So my Weil-McLain WGO-5's, measured by our boiler techs at 82-85% combustion efficiency, are really only 75%-78% combustion efficient because the boiler tech's combustion analyzer does NOT measure the latent heat of vaporization loss, which is about 7% for oil combustion. And as you know, natural gas combustion produces more water vapor than oil combustion, so when the boiler tech measures 83% dry gas combustion efficiency in a gas boiler, his dry gas analyzer is missing the 15% or so latent heat that is ALSO lost in that combustion product stream.

What confuses people is that, at 85% "dry gas" efficiency, the 15% of BTU's being lost in the dry gas stream happens to be roughly equal to the 15% ADDITIONAL BTU's also being lost as latent heat of vapor in natural gas combustion products.

I have verified this myself by taking the combustion gas parameters that our boiler tech has measured, and applying them in the so-called "dry gas" heat loss equation. And the result agrees with the 82% (most recent) combustion efficiency measurement by our tech. And I can tell you categorically that the "dry gas" equation that gives the "correct" 82% result does NOT have any factor in it for latent heat of vapor loss. That is literally why it's called the "dry gas" equation.

@Captainco has correctly harped on this in past threads on this forum, which I will not attempt to resurrect. But the bottom line is that any combustion efficiency measurement taken with a standard combustion analyzer used by most boiler techs is a "dry gas only" heat loss that does NOT account for the additionl 7%-15% latent heat of vapor loss (depending on fuel) in the combustion stream when the flue gases are not condensing.

What further confuses the issue is that once the combustion gases have cooled sufficiently to raise the "dry gas" efficiency measurement to above 89% or so, the vapor does begin condensing, and then that latent heat IS recovered. So when you get to "dry gas" combustion efficiency measurements of 90% and above, at which point you may be condensing most or all of the vapor, THEN the "dry gas" combustion efficiency number is closer to accurate, because the latent heat of vapor IS begin recovered (even though the "dry gas" equation doesn't actually "know" that or account for it.)

So it is correct to say that the "dry gas" efficiency measurements in the 90's or above DO get closer to being correct, because at that point the vapor is condensing and most of that vapor heat is being recovered. But when the "dry gas" efficiency measurements are below the condensation temp of the flue gas, that latent heat of vapor is being lost IN ADDITION to the "dry gas" losses measured by the combustion analyzer. Which is why, for non-condensing combustion flue gases, the latent heat of vapor loss must be ADDED to the dry gas loss to get a correct number for the total heat lost in the flue gas stream.

Jamie Hall

I totally agree that the combustion measurements taken by the equipment used in the field are dry gas measurements, and make no allowance for the water vapour in the combustion gas.

I was — and remain — much more concerned about the problem which is wide spread that the difference in higher heating value and lower heating value of the fuels is not understood — even when people are aware of it — and that recovery of that latent heat is the key to the high apparent efficiencies of condensing boilers.

From my point of view I take any efficiency number with a very large "oh really?" attitude — just as I take things like horsepower ratings or other power output ratings. Provided that various devices are rated using exactly the same technology and procedures, those ratings can be of value as a comparison — but nothing much more than that.

Heating systems are particularly problematic, of course, since overall efficiency of the complete system is hard to define, never mind measure. It's a lot easier with power machinery — as an ME, you would know that!

jesmed1

In a crude attempt to illustrate the difference between "dry gas" losses and latent heat of vapor losses, here's a rough graph showing combined "dry gas" and latent heat of vapor losses in natural gas combustion.

What confuses people is that an 85% efficient (as measured by a typical "dry gas" combution analyser) natural gas boiler is losing 15% of its BTU's as "dry gas" heat PLUS about 15% of its BTU's as latent heat of vapor loss. So an "85% combustion efficient" natural gas boiler is actually losing 30% of its BTU's, half as dry gas and half as latent heat of vapor.

As flue gas temp decreases and the "dry gas" efficiency increases, the vapor begins to condense, and that latent heat begins to be recovered. But the "dry gas" combustion analyzer doesn't know that. It only "knows" what the dry gas is doing. And as the dry gas efficiency increases above 90%, more vapor condenses and more latent heat is recovered.

The shape of my latent heat of vapor curve may not be entirely accurate, but the general trend is correct. Roughly 15% of BTU's are being lost as latent heat, in addition to the dry gas losses, until the vapor begins to condense. And when the flue gas temp gets low enough, in theory, you recoved 100% of BTU's from both the dry gas and the vapor.

I apologize for sidetracking this thread, but this subject of latent heat of vapor loss confuses so many people here that I think it's worth trying to de-mystify.

Jamie Hall

I think we're pretty much on the same page, @jesmed1 . Different terminology, perhaps.

In reality, it really doesn't matter that much — until the numbers get baked into regulations or specifications. Then it starts to matter… a lot.

TommyA

Thanks everyone. Maybe I can summarize like this:

GPM-ish = Btuh-ish / (500 * deltaT-ish)

No wonder the old boiler was 3X over sized 😂

Jamie Hall

One more poke at the bear. What matters, as I noted, is system efficiency, not some instrumental measurement (which, however, is very useful to make sure the tuneup is correct). So what really matters for heating is BTUh output vs. energy per hour (gallons per hour, cubic feet or therms per hour, kiloWatts per hour, whatever) input.

Fortunately, not that hard to measure…

Edit: Just ran that test on Cedric. 86% of the higher heating value appears as useful (steam) output. Not too shabby — thank you, @Charlie from wmass !

Wellness

@HotRod What software is that you have in the screen shots? Caleffi's?

hot_rod

Hydronic Design Software

Find it a www.hydronicpros.com

There is a free demo version, it is Windows based only

The pump database is being updated also

Wellness

@hot_rod. Thanks.