We have been using ISE for ages (2000) to test for ammonia and need to find out when it was approved for this use. Have a regulator who says it was only approved in 2010.

mark

Mark:

Ammonias by ISE were approved awhile back with one caveat: you had to distill the sample then use ISE for direct measurement.  Unless you could provide proof that distillation was not necessary.  This meant a correlation study-- running samples that were not distilled vs samples that were distilled.  If no difference then there was no need for distillation.  The same thing for fluorides.  Inspector might have been alluding to this facet of the method.

Mark,

Who's your permit manager?  You need to contact somebody at the DEQ lab and have them contact him/her directly and straighten them out.

As an aside, if possible I'd switch to a different method.  ISE has become the go-to method because it's relatively cheap and easy; but the method has a serious problem with drift, especially at low concentrations.  How low is low?  ASTM sets the lower limit at 0.5 ppm, which is much more realistic that SM's 0.03.  SM's own data (Table 4500-NH3:I on page 4-112 of the 21st Edition) shows horrendous loss of linearity at 0.1 ppm and shows poor effluent precision at 0.8 ppm (+/- 38% !!).

There was a long thread on the "old" WEF lab forum about all of this (unfortunately now lost because WEF redid their online forums).  The problem with the ISE is that the measurement is diffusion limited, NOT kinetic limited.  Thus, anything in the sample that would interfere with the diffusion process will mess up the electrode performance and give you crappy results.  Thus, the necessity for distillation.  If you want to really tune up your ISE method, run a series of standards of decreasing concentration and find out where you start to lose linearity.  I have a hand-drawn calibration plot I use to illustrate what happens, and I can scan it and send it to you as a pdf.  The bottom line is, if you try to quantitate in the nonlinear region, you'll overestimate ammonia, big time.  If you look at the table in SM I referenced above, you'll see what happens:  you get spiked blank recoveries that are huge, AND the recoveries bounce all over the place.

Okay, enough of the soap box rant.  I just think that ISE ammonia is a case of cheap/easy not being the best way to go.

Chuck Lytle,
City of Portland, OR

Mark,

We currently use an OI/Perstorp flow injection analyzer.  We're in the process of buying a discrete analyzer, but of the five vendors we looked at, none could get down to where our permits require us to go.

The nice thing about the FIA is that you don't have to distill first.  You can if you want to, but that would only be if the data showed you had an interference problem.  We're lucky enough to not have to do it.

Our low-level MRL for ammonia is 0.01 ppm as nitrogen.

Chuck Lytle

Hello

What brand of ISE do you use for NH4 analysis?

Thanks

R

You may want to consider to do distillation and then titration of your sample. This method works better, much much better recovery on standards and spikes.

The electrode needs lots of attention and since you're loosing gas while measuring ammonia so the readings will be jumping a lot.

The titration method is an old techniques but works perfect. You may may want to consider gettiing a automated distillationa and also a auto-buret to help out  if you have the buget.