Comparing Results to Outside Labs

[BiodieselGC]

Hey everyone,

I'm new to running a lab and as much as I learned in school, there was little focus on troubleshooting specific instruments. Hopefully someone with more expertise and experience can point me in the right direction.

Here's my main issue: I obtain consistent results from sample to sample with my GC. If I perform 2 runs with the same sample, I essentially get the same results. We sent samples to 2 outside labs last week and there was a considerable discrepancy between their results. My results matched those of one lab, and were far from the results of the other. We don't send samples to outside labs often, but the few times we have, I have only obtained consistent results with this one last week. I would assume since I am more often unable to confirm results with outside labs my results are off.

I was not here when the GC was calibrated and my gut feeling is it needs to be recalibrated. However, I don't want to waste money on the standards if this is not necessary. The areas of my internal standards are consistent, so I don't believe they are the issue. I don't believe there is a leak, either. Does anyone have any suggestions of when to check or look for?

Thanks.

[GasMan]

Which biodiesel methods are you using?

Gasman

[BiodieselGC]

Are you asking ASTM or EN Standards? If so, ASTM.

[Don_Hilton]

Suggestion: Get some standard mixes and run them ocasionally. This will be a check - and unlike sending to an outside lab, you have a known sample with a known answer. These may cost a bit, but I expect that there is a cost to wrong answers also.

[Muhammad Akbar Khan]

Dear, you must run qc sample first, if it gives you results out of qc sample, then you need to recalibrate it. Always make sure your gc is working properly before sending outside,

Regards
Akbar

[Johnny Rod]

If you can make samples and obtain broadly consistent results, it suggest the problem lies more in the calculations (interpretation), I think. How do you calibrate the instrument? Do you not run a standard with each set of samples, or at least each day?

Unfortunately lab services offered aren't always the benchmark they should be. We take part in some correlation studies and it amazes me what people think are reportable results, even when they know who they are reporting them to. 100.6% anybody? (plus/minus very little)

[chromatographer1]

BiodieselGC wrote:

" I'm new to running a lab ..... Hopefully someone with more expertise and experience can point me in the right direction."

" Here's my main issue: "

" I obtain consistent results .... We sent samples to 2 outside labs last week and there was a considerable discrepancy between their results. "

" .... I would assume since I am more often unable to confirm results with outside labs my results are off. "

WHY ?

" I was not here when the GC was calibrated and my gut feeling is it needs to be recalibrated. "

GOODNESS ! HOW LONG HAVE YOU NOT RECALIBRATED USING A STD MIX ? YOU ARE NEW TO LAB WORK ! TIME TO LEARN ABOUT GLP !

" However, I don't want to waste money on the standards if this is not necessary. "

STDS ARE NOT A WASTE OF MONEY. THEY ARE ESSENTIAL TO PROPER OPERATIONS.

" The areas of my internal standards are consistent, so I don't believe they are the issue. I don't believe there is a leak, either. Does anyone have any suggestions of when to check or look for? "

YEP ! RUN CHECK RUNS USING A STD MIX. YOUR OBSERVATIONS ABOUT CONSISTENT AREAS SHOW YOU ALREADY HAVE A GOOD QUICK CHECK ON YOUR RESULTS! GOOD JOB !

The fact that one outside lab was wrong and one MAY have been right (the one that agreed with you) why would you believe one over the other? Verify your results with a std (I know it costs money but so does your labor charges and the cost of heating and lighting the lab, right?)

Do it right and you won't have to defend yourself or your results. USE STDS.

and keep up the good work.

Rod

[chadac]

First of all, what are your analytes and the detectors you are using? Secondly, I'm going to have to reiterate what everyone here is saying. You need to have standards and recalibrate daily. I don't know what type of work you're doing, but the general idea you want to go for is,

A) Having an exact match standard made of what you're expecting your product to be.
B) Having a series of standards made to try to achieve linearity, ex. a 5 ppm standard - 25 ppm - 50 ppm - 100 ppm, so on and so forth.
C) Having a standard that is no less than half of your expected concentration and no more than twice the value you're expecting.

Options A and B are my main recommendations, depending on what your application is one can be advantageous to another. Ex. being if I have a sample that contains 10+ analytes ranging from percent level (mol) to ppm level, I'm going to need an exact replica standard to get the best results, especially in the occasion where I'm expecting dual peaks.

If you're being lazy, you can get away without recalibration with certain analytes somedays, for example, I know my cal tables with say, methane and ethane will reproduce near identical response factors day in and day out, but any tables containing acetylene will absolutely vary from day to day, even down to sulfur species which will drift in a matter of half an hour in my experience. Regardless, you're going to need standards, to not have them is very risky to your operation and will have you raked over the coals in the event of an audit from any customers, who will want to see the paperwork on your standards along with calibration records.