The effect of SCR on design of flue gas systems

[athomas236]

Gentlemen,

I am just beginning to get involved in a project that involves installing SCR on an existing coal fired utility boiler.

One of the concerns is the impact of the installation of SCR on the design of the furnace buckstays and ducts. At this time we have little information about the existing system, but will be trying to track this down over the next few months.

It has been suggested that we contract with someone to develop a dynamic model of the flue gas system, validate this against the existing system and then use the validated model to predict pressure variations after SCR has been installed and identify control measures that could limit such variations. If the pressure variations are outside the design of the existing system then the predicted pressures will be used to redesign the furnace buckstays and ducts.

My problem is that I am uncomfortable with using clever dynamic models for design and wonder if anyone has experience with such models that would show they have been used successfuly on similar projects.

Best regards,

athomas236

 

[pmover]

not that i'm aware of . . . but a suggestion is to adopt/implement only proven systems/designs based on proven models. that is unless there are reputable companies/individuals willing to put forth the time & effort, risk and responsibility, etc. in such a model & design.

not much to work with, but hope this helps.
-pmover

 

[rmw]

Athomas,

How would you have approached this back in the days before fancy models?

Will the model builders know as much about the system as your experience will tell you?

rmw

 

[metengr]

athomas236;
We have gone down this path as an owner. Your best approach is to partner with a boiler OEM, preferably the OEM of your boiler, and work in cooperation regarding design, procure and construct. If you do not work with the specific boiler OEM, I would suggest teaming up with another boiler OEM that has the breadth and depth for environmental controls.

 

[davefitz]

I guess the short answer is that if you do contract with a 3rd party to dynamically model the furnace + flues + fans, ask to review their past models for other clients to assure yourself that their model can accurately predict the future performance after installation of the scr + more powerful ID fan + faster dampers. Their past model's prediction of a predicted MFT transient vs field test data after SCR conversion of a real MFT is the case to review.

The models are sufficiently complex that they impress the uninitiated , espescially if they supply colorful curves of the transient response. But they remain fully fudged, using field data from the 1970's to estimate the rate of flame collapse, and further "adjusting" the rate of decrease of coal fed to the furnace following MFT based on data from a past MFT of the unit. One trick one party had used to fudge the result was to artificially slow down the rate of closure of the BSO burner shutoff damper- be sure not to interpret this as a permissive to actually slow down the speed of these dampers- this would become an avenue for a furnace puff, and the need to rebuild the furnace.

 

[boilerone]

athomas236,

Let me recap your question and hope this is what you are asking for: you are worrying about the changes of pressures both upstream and downstream of the SCR unit after it is added because of the pressure drop and its variations created by the SCR.

Typically, you'll try to maintain the upstream pressure not changed thus you won't have to mess with a majority of the existing system including furnace MFT setting, etc. which are tied to the allowable furnace pressure variation range. Only downstream pressure and dynamics are changed because you'll have to increase the ID fan vacuum pressure, or capacity to take care of the pressure drop across the SCR plus the effect of ammonia injection. For this method, the buckstays on the furnace side need not to be worried about but the ductwork stiffeners (plus any component all the way to the ID fan) downstream of the SCR need to be checked for sure because of the higher negative pressure.

I would not suggest trying to even out the pressure changes across the SCR which means to allow the furnace side pressure to go up a bit and pressure downstream SCR to go down a little bit therefore it requires only increasing ID fan vacuum for 1/2 of the SCR pressure drop. By doing so, it not only will reduce the gas flow velocity thus affects boiler performance/output, the furnace can easily go into positive pressure and start to leak out flames and sparks from many places. A furnace is normally design as always under slight but constant negative pressures for two main reasons:
1. safety, because it is very hard to make sure all structural seals are perfect plus many places might crack open after a few years. Neg. pressure can make sure only outside air be sucked into the furnace and no furnace gas can get out to make the boiler smoking.
2. if furnace pressure is near ambient then pressure fluctuations from combustion and the FD and ID fan pressure changes can make the furnace pressure to vary beteen - and + to make the furnace walls to go into a visible 'in and out' kind of pulsation movements (as if it is breathing) which is dangerous because it might cause fatigue damages of the buckstay connections, seals and furnace-attached pipings, etc.

Regarding modeling the flow with a computer, I don't think it is necessary for the pressure dynamic change considerations because the SCR vendors typically supply guaranteed pressure drop values for your easy calculations for the worst situations or different 'cases'. But if it is for the SCR and related ducting design/change themselves, YES, it is normally required to make a CFD finite element model to show how even the gas and ammonia flow after passing the AIG and enters the catalysts so as to get the most economical layout and highest NOx reduction efficiency. The CFD is a proven method and one of the most popular tools and must-haves for the OEM suppliers. I would question any one who does not have this gadget when he/she tries to sell you SCR systems.

Even though not necessary but, 'for convenience', a good SCR supplier should be able to use its CFD model to predict the pressure dynamics for you, for free or as part of the contract. These models SHOULD be trusted.

Boilerone

 

[athomas236]

Gentlmen,

Thank you all for your advice which I will review tomorrow. We have a telephone conference with our client tomorrow; I will keep you informed.

Best regards,

athomas236

 

[athomas236]

Gentlemen,

Sorry for delay but the meeting I expected has been postponed twice so not sure what the client's next move will be.

With regard to your comments I reply as follows:

(a) rmw
Back in the late 1970s when I was a boiler designer, I think we used a standard pressure of +/-10ins WG but after we had three "implosion" incidents on one plant this was reconsidered. Cannot remember what the outcome was.

The incidents were not really "implosions" as such just operators closing the burner air slides during start up leaving the furnaces subject to full ID fan shut off head. We spent weeks redesigning ducts and buckstays for the shut off head but the client was put off by the costs.

Since becomming a consultant, I have always specified NFPA.

(b) metengr
The intention is to team up with an OEM although not the original OEM, they no longer exist. As you can imagine, I have reservations about using dynamic models as a basis of structural design.

(c) davefitz
Based on your suggestion, I have asked for details of previous successful experience but neither our project team or the client seem unwilling to pass on this request.

(d) boilerone
The problem we have is that the client doesnot have any information regarding the original design pressures of the existing plant which is more 40 years old. As a consequence, it is not possible to assess the impact of any change in steady or dynamic pressures during an MFT.

My preference is to use NFPA guide of +/-35ins WG.

Best regards and thanks again,

athomas236

 

[rmw]

athomas236,

I saw one coal fired utility put actuators on their ID fan dampers that would literally rip the fan out of the ground just so it would act fast enough on a burner or mill trip and isolate the ID Fan.

The plant engineer said he didn't want any excuses that the ID fan dampers didn't close and sucked the boiler walls in.

I think he was trying to react to the "bubba" factor too and make it so that if idiots were operating the unit, the "implosion" wouldn't be the instrument departments fault (as they were getting blamed for damage caused by operator errors.)

rmw