oxidation catalysts- formaldehyde & HAP

[davefitz]

The recent 5 Mar 04 EPA rule (40 CFR part63) will require all new staionary combustion gas turbines to add oxidation catalysts , to ensure the outlet formaldehyde is below 91 ppb.

In the case where we currently have an existing deNox SCR in the HRSG ( downstream of gas turbine) , there is space to add an oxidation catalyst , nominally called a CO catalysts. The spool piece which was provided to allow addition of a new CO catalyst is upstream of the deNox SCR .

I heard that the oxidation catalyst will also convert NO to NO2, and if the NO2 exceeds 50% of NOx, the downstream deNox catalyst will lose effectiveness.

Can anyone comment on the accuracy of this latter statement, and if there are methods that can be used to preserve teh activity of the downstream SCR

 

[PeterAB]

Davefitz:

sorry that I'm not being more helpful, but your posting prompts me to ask another question: can anyone out there measure formaldehyde concentrations down to 91 ppb with any kind of repeatability? How!? It's news to me.

 

[bchoate]

bchoate
EPA method EPA 8315A tests for formaldehyde in air, liquids, and solids by derivatizing with dintrophenylhydrazine and applying gradient HPLC. Air samples are absorbed onto a cartridge which is later extracted for the EPA test. The method can detect formaldehyde down to 6 ppb.

 

[hrsgguru]

Davefitz:

You'll probably find that most SCR catalyst vendors base their guarantees on a 20:1 NO to NO2 ratio (i.e., NO2 is 5% or less). This is common for most GT exhausts without oxidation catalysts. The addition of the oxidation catalyst (OC) will definitely affect that ratio, but the extent will be determined by the specific OC material, the actual ratio of NO to NO2, the exhaust flow distribution over the catalysts, and most importantly the operating temperature of the OC. OCs are usually more effective at higher temperatures (400 to 500°C) while the SCR catalysts are more effective between 300 to 400°C. Thus, your OC installation directly ahead of your SCR system will probably place it in the lower temp zone and may result in lower NO to NO2 conversion.

This problem has only recently been observed (over the last 6 to 8 months) and is starting to be addressed by the vendors. You'll need to get both your OC vendor and your SCR vendor together to see how the two will interact. In the worst case, you may have to add volume to your SCR catalyst. In the best case, you might not have to do anything.

If you have any appreciable amounts of sulfur in your fuel, you may now encounter another problem due to the OC addition. Fuel bound sulfur leaves the GT as SO and SO2. Both the OC and SCR will convert these to SO3. This will in turn react with the ammonia from the SCR to form ammonium bisulfite. And this substance will condense on your tube surfaces in the back of the unit and greatly reduce your extended surface effectiveness over time. But there are ways to deal with this if it should occur....

 

[cub3bead]

Adding to hrsgguru's last paragraph.

Ammonium bisulfite will precipitate as a white powdery deposit on the extended surfaces in the HRSG backend (at T<300 deg F). If the ammonium bisulfite gets wet it readily forms a very acidic solution that will attack the tubes and fins. Duplex stainless is one of few materials suitable for replacement surface in this case and it will cost big $$$.

 

[davefitz]

WE normally fire natural gas , with odorant. The sulfur content of the odorant is low enough that ammonium bisulphate from oxidation of SO2 to SO3 + ammonia slip is not an issue.

The liquid fuel in the tanks is ultra low kerosene. After Sept 2005, there is to be available in the US ultralow no 2 fuel oil ( less than 6 ppm S), so we may stock up on that to forestall any issues related to sulpfur.

 

[bonzoboy]

If you have a gas turbine, you probably have a Low NOx combustor. So your NOx emissions would be low to begin with. I recall that the SCR actually works better if the NO2 and NO are near equal amounts. Also, the SCR you have installed is probably oversized to begin with; so the use of a catalyst (oxidation) probably won't have much of an affect (the ammonia consumption change might only be in the noise of the measurement)

 

[davefitz]

bonzoboy:
Correct. As per GS Madia , the equimolar reaction with NO + NO2 + 2NH3 +O2 is much faster across the catalyst, so as long as the NO2 concentration does not exceed the NO concentration , the SCR will appear too be oversized. Once the NO2 levels exceed the NO levels, though, the opposite occurs, as the NO2 reaction is much slower than the NO reaction.

The oversizing of the SCR catalyst is not a permanent feature; it is originally oversized to account for aging and poisoning effects- it will eventually need to be replaced as the poisoning / aging reduces its reactivity. The initial oversizing is what determines the replacement interval.

 

[bonzoboy]

NO-NO2 conversion: The conversion rate for NO-NO2 on an oxidation catalyst is reported to be VERY slow. 1000x slower than the oxidation of CO to CO2.

Also, thermodynamics will tell you that the higher the gas temperature, the lower the NO2 portion.