Radiative vs Convective Boiler HT

[bonzoboy]

A conventional D-type boiler I looked at appears optimized for substantial radiative heat transfer (based on the wall tube surface area).

If, instead of direct firing with natural gas, this same type of boiler is placed in a waste heat recovery operation (where there would be much less flame radiation), how much might performance on the boiler be degraded? (performance measured by lb/hr of hot water at fixed temperature, or overall thermal efficiency).

In a waste heat (HRSG) application, do the boile manufacturers increase the convective heat transfer tubing to make up for the lower radiative transfer?

Bonzoboy

 

[athomas236]

In unfired HRSGs the radiation heat transfer comes from the CO2 and H2O in the flue gas and also is dependent on the tube and fin dimensions. Usually radiation heat transfer in such boilers is so small it can be ignored.

In fired HRSGs the flame itself is clearly the source of radiation heat transfer. The amount of such heat transfer depends on the fuel burned and can be estimated using Stefan Boltzman equation. For gas I would use an emissivity of about 0.5 and for oil a value of about 0.85.

This radiation will will affect only the first one or maybe two tube rows of the first bank of heating surface

 

[bonzoboy]

Thanks. But assuming that I make no changes to the design of a boiler that is normally fired, and instead I put it in a waste heat application, I assume that the amount of steam or hot water I can generate (with no changes to the heat transfer area) is decreased. But by how much?

Any thoughts.

 

[rmw]

Here is an approach that might get you close to where you want to be. I really don't know if it is valid or not frankly, but it is where I would start (short of consulting the OEM of the boiler) to get a rough estimation of whether or not the boiler in question will work as you want it to.

Assume that the amount of steam generated per tube in the boiler is a constant. In other words, assume each furnace wall tube generates more or less the same amount of steam per tube as a generating tube would for the purpose of this analysis. Order of magnitude wise it has to be in the same neighborhood.

I think this is a fairly safe assumption, because once nucleate boiling starts in the tubes, whether they are furnace tubes being heated with radiation or generating bank tubes being heated by convection, certain limitations with regard to bubble velocity and water flow in the tubing exist within certain ranges.

Deduct the furnace tube count from the total tube count, and take the ratio of the steam production of the generating bank to the total number of tubes in the boiler as your derate ratio.

In other words, if you have 200 tubes and 50 of them are furnace tubes, (either side of the "D" wall) then your derate is 25%. Your furnace tubes would of course contribute to your overall heat transfer in the case of the waste gases, but we are going to ignore their contribution for now.

This, of course, assumes that your waste heat flue gas temperature at the entrance to the boiler furnace would be the same as the theorhetical temperature of normal flue gasses at the furnace exit/generating bank entrance when firing a fossil fuel like gas or oil. If not, then you have to take that ratio into account as well. And, if it is the case that the temperatures are substantially different, to my way of thinking, this little quick check does not work.

It's a stretch, but other ideas don't seem to be coming forth, and I once had a situation much like yours where direct radiation from the furnace was missing because the furnace was remote from the boiler and it was a real problem with respect to the derate that we had to take because we did not have the convective heat transfer surface area to make up the difference that we lacked from the radiation of the furnace.

To answer another of your questions, you will notice that in HRSG tube banks, the tubing is finned in order to maximize the convective heat transfer.

rmw

 

[bonzoboy]

Thanks RMW. That's almost the kind of problem I'm trying to resolve. I'll give that a crank....