filters

Triangle NoTES - December 2012

Merry Christmas, a Happy New Year, and a safe holiday season to everyone!

This newsletter is to give an insight into one of the impacts of using the Method 5 type filter in the Method 25 train during inlet monitoring.

While this effect has not been seen in a large number of tests, it has been indicated in some and clearly shown in others. The concern has been giving the information on those tests without possibly identifying the source while there are still possible legal issues to be reconciled. I believe I have gathered enough information to show the effect without compromise of any identities.

I have combined the information from a couple of sources which have some similarities and a good record of data. These have multiple inlets and a single outlet from a control device. I have mixed them in such a manner that the effect can be seen without knowing which inlets are connected to which outlets, in order to prevent identification of the sources.

I have taken a ten year series of data, using the last year as a comparison to the previous nine year average concentrations. As the averages indicate, there has been little change over the period until the last year when the filter was changed and the size of the aerosols sampled was limited. The results are not corrected for any flows and are rounded ppmC concentrations.

 

Source

Nine Year Average

Last Year

Change

%RSD

Inlet One

4410

4540

+3%

13.5%

Inlet Two

4590

4770

+4%

10.1%

Inlet Three

1170

55

-95%

17.6%

Inlet Four

1900

180

-91%

11.9%

Inlet Five

760

27

-96%

10.6%

Outlet One

220

215

+2%

18%

Outlet Two

195

195

+0%

12.6%

 

 

As can be seen in the two outlet and the first two inlet concentrations, there is not a lot of change from the long term average. The change is in the other three inlets which are all similar in nature. The 90% minimum drop in the inlet loading compared to the long term average shows the problem. The destruction efficiency of the control device dropped because of the loss in inlet loading, which was removed from the sample by the filter, but not from the stream to the control device.

All of the reported initial concentrations and all but one of the reported last year concentrations exceed the levels for which Method 25-A would be considered as an option. This would be further compounded by the lack of filtration required in Method 25-A, which should in reality place the concentration much closer to the nine year average. The use of Method 25-A in the projects where the Method 5 type filter in the Method 25 sample console has apparently caused a drop in loading indicated a much higher concentration than reported by those Method 25 analyses. The use of Method 25-A without filtration has a calculated control device efficiency approaching a more realistic and believable range than Method 25 with the new filter. Generally the Method 25 analysis reports a higher concentration due to the lack of the response factor differences between the sample gas and the compound used for calibration in Method 25-A.

Of course, there could always be a modification to the new filter specification used for Method 25, but that would generally be done on a case-by-case basis and after a first attempt had shown the problem affecting compliance. The cost of that first failed test would have to be written off by someone, but hopefully would result in sufficient data to allow for future modifications to be accepted for that source.

One of the issues facing this type of source is the filtration of what appears to be aerosols rather than solid particulate matter. This means at some point during the process the aerosol will become a vapor and will have to be dealt with as such. It could possibly be corrected by sampling at another location closer to the control device, if possible, where this transformation had already occurred. If not, removing the aerosol as a particulate would seem to require some type of additional control device for the organic compounds. Any additional series of controls are to be unnecessary for actual control based on historic data, but might be necessary for the definitions to be used.

I suppose this is why so many people dislike the way the regulations are handled, because it appears it could cause unnecessary costs to be incurred. I am not sure that would actually be the case, but there is the appearance.

I believe when a known problem exists and can be corrected by merely not making a change to the inlet filtration sampling the reasonable course is to do just that. Making a change to the inlet sampling that really has no effect, since the majority of the filters used previously would still be used, and the sampling is still within the process seems counterproductive. I can see where there would be an issue with an emission to the atmosphere, but not within the process itself. Thus, the correction to the outlet requirement is understandable, but not so with the inlet filter given there will be issues created with future testing and will impact the comparison to previous data.

Wayne Stollings

Triangle Environmental Services, Inc.

Wstollings@aol.com

 

Triangle NoTES - June 2012

This newsletter is to give notice of our options for filtration for the Method 25 sampling. Starting July 1 if there is a possibility of aerosols in the inlet source to be sampled, you need to let us know so we can ensure the special INLET ONLY filter will be installed in the appropriate train. Do not use the wrong train for the inlets and outlets as it can double the negative bias potential for your source.

To bring things up to date, there was a project in which the reported concentrations for both Method 25-A and Method 25 using a less restrictive rated glass fiber filter showed much higher concentrations than the same tests using the glass fiber filter specified for Method 5. This information was brought to the attention of the EMC since there was no specification for the Method 25 glass fiber filter retention. The concept had been to use a similar filter and temperature division point as Method 5 so there would be little chance of duplication of reporting as particulates and organics. Thus, the EMC explained, they would be changing the specifications in Method 25 to require the similar filter criteria as was in Method 5. This seemed logical and we made the change as EMC had indicated was their intent. The potential problem with inlet sources to control devices was overlooked by everyone in the discussion, including myself. This was fairly quickly made evident when we subsequently reported inlet concentrations well below outlet concentrations to a control device. Where there were comparable Method 25-A results, the inlets for 25-A were closer to the expectation and the outlet concentrations were similar to the Method 25. As a result, we have again contacted the EMC to have them consider the change of filtration to Method 5 equivalent for only outlet sources. I believe the inlet to control devices should have minimal filtration requirements because the aerosols and particulates going into the control device are all part of the inlet carbon loading for the calculation of destruction efficiency. Any removal of reported carbon from the inlet sample, but not the inlet loading would bias the results against the source. In some of the cases, this negative bias is very significant.

We recently reported on a project that again raised the concern over filtration impacts on measured inlet concentrations. This project is a good reference point because it has been tested repeatedly, been stable each test cycle, and there have been no significant changes since the last test cycle. This type of long term historical data can really highlight the effect of a relatively small change.

This is a facility with eight sources being sampled each test cycle. The results for this cycle were very similar to the results of last cycle, with the exception of the three sources where there was a very high probability of aerosols. The reported concentration for these three sources this cycle approached ~2% of the reported concentration last year. This would have been great if it had been a destruction of the TGNMOC, but that was not the case. The outlet sources were in line with the same concentration seen in previous cycles. That indicated there probably was the same general concentrations in the sources or that the efficiency was off just the amount of the loss in the aerosol sources. The coincidental efficiency change is not very likely to have happened, especially with the other projects where similar effects have been documented in various and different ways.

In short, this facility is tested about the same time every cycle by the same company. The results for the different sources are similar between tests performed on different cycles, except for the three sources this year. Our equipment and procedures are the same as always, except for the micron rating for the heated filter, which was only changed this test cycle.

We currently use a filter from the same manufacturer as we did for all of the previous projects. It is listed as a binderless glass fiber filter, just as was the previous filter we used. The only difference I can determine is the size of the particles which are supposed to pass through the filter. We had previously used the GD120 filter until I was told there would be a change to using the Method 5 specifications which the GC50 style should meet.

The Method 5 specifications are different from those listed here, which is somewhat confusing, since the specifications for water or liquid are different than air.

Since there has been no legal change to the method as of yet, this situation has caused us to make the following changes to our procedures. Starting in July, any sample consoles with the lesser filtration capacity we will use in INLET sampling will be specially tagged and the replacement filters for it will also be tagged as INLET only, unless and until there is some official change to the method to cause another change.

This filter will be provided for INLET sampling as was the historical case with our equipment.

GRADE GD120- Course porosity, fast flow rate, with a 0.9Fm size particle retention. Used as a pre-filter (1.2Fm or smaller). High wet strength and good dirt holding capacity. Binderless glass microfiber. Comparable to Grade D or Millipore APFD

This filter will be used for OUTLET sampling as was intended by the EMC guidance to us.

GRADE GC50- Fine porosity, medium/slow flow rate, with a 0.5Fm size particle retention. Used as a pre-filter (0.45Fm or smaller) and suspended solids analysis of industrial waters and wastewater. Binderless glass microfiber. This filter is comparable to the Grade C or 934-AH®

The other filter specifications:

934-AH - Fine porosity, fast flow rate, with a 1.5Fm size particle retention. This material is the standard for suspended particle analysis in water and cell harvesting. RIA scintillation counting. Binderless borosilicate glass microfiber.

Grade C - Fine porosity, fast flow rate, with a 1.2Fm size particle retention. Primarily used for RIA procedures and harvesting lymphocytes. Binderless borosilicate glass microfiber. Grade C has several sub groups

Grade D - Coarse porosity, fast flow rate, with a 2.7Fm size particle retention. Well suited for higher volume and repetitive laboratory filtering. Also ideally used as a general pre-filter to extend primary filter life. Binderless borosilicate glass microfiber.
 

Wayne Stollings

Triangle Environmental Services, Inc.

Wstollings@aol.com