WEF Discussion Forums
Laboratory Management and Technical Issues
For those who run the COD test on raw sewage for regulatory purposes, would you be willing to share your SOPs and/or your quality control measures that you run with each batch? Thanks! firstname.lastname@example.org
It really all depends on what your regulatory agency requires.
We run Hach 8000 and are required to establish our own curve on the spectrophotometer(we do this quarterly with standards at our reporting limit, midpoint of the range, and top of the range - 20, 800 and 1500 mg/L).
We perform MDL's annually per 40CFR136.
Daily we prepare a reagent blank and a midpoint check (800 mg/L). We analyze the reagent blank and midpoint check immediately after zeroing with a instrument blank. We also reread the midpoint check after every 10 samples and at the end of the analysis.
We also run a duplicate with every 10 samples.
In addition to what Gayle mentions, we are also required to run an known check standard from a source other than that used to make the calibration curve. We also run spike samples every 10 or fraction of 10 samples. Our recovery requirements are ±5% for the standards, midpoint and second source check, and ±10% for the spike sample recoveries. For our curve to be valid we require a R value >0.995.
Gayle, just out of curiosity, what kind of recoveries do you get on your 20 standard? We had to go to two different levels of curve because we weren't getting precise enough recovery at 20. We use the regular COD vials for values greater than 100 and then we have vials made for a range of 0-150 that we use for values lower than 100.
Sorry, I actually mis-typed, we are actually running at 50 mg/L and we are using the regular COD vials. We are using criteria of 50-150% recovery, but generally run between 70-130%. The criteria we use for the 800 mg/L standard is 90-110% recovery, but generally get from 95-105%.
For the curve, we require the r2 value to be >/= 0.995. Our daily curve checks at 50 and 800 mg/L are from a source other that the one we use for the curve standards.
DSmith, how are you doing spiking? We considered this, but could not work out a good way to do it with some of the sample volumes we receive (we also use the same sample for BOD, TSS and TDS analysis). Fortunately we have not been required to do spikes by our regulator.
I'll try to be more careful with my typing next time.
Sorry, the anonymous post is actually me, Gayle. I forgot to logon before posting.
We make a 40,000 mg COD/L solution to spike with. 8.50 g potassium acid phthalate dissolved in reagent water to 250 mL will give you this concentration. Add 0.5 mL of that to 49.5 mL of sample will give you a spike concentration of 400 mg/L.
Yeah, a 50 is a bit more reasonable than 20. That % recovery sounds like what we get too.
Thanks for all the replies on this subject. dschultz
We only run the COD analysis for our own use but we allow industries to do a correlation study BOD vs COD to calculate BOD for our user charge. So we need to be able to analyse COD accurately and be able to get comparable results with the industries for their discharge to our sanitary sewers.
We use Hach 8000 method but we do not a standard curve. We use the preprogramed curve of the Hach method. We do run KHP COD standards and have acceptble ranges for these standards. Hach has 300, 800, and 1000 mg/L COD standards. Lab Chem INC has a 300 and 1000 mg/L standard, while North Central Laboratories have a 500 and 1000 mg/L COD standard. These are from different suppliers so they could be used as an independ standard if you generate a calibrtion curve.
For the low range Hach COD method we make a 1 to 2 dilution of the 300 mg/L standard, 150 mg/L, and check the recovery with that method.
We use Hach 8000 method for industrial effluent control, but for legal considerations we need to accredit the method by ISO 17025. We are looking for a reference method wich mention the HACH 8000, to avoid doing the complete validation of the methods. We wanted to ask you if somebody did this validation or if you know a reference methods that incorporates the HACH 8000 as reference methods.
Excuse me for my english