Question for the scientist/geeky types
 

Proximate Analysis: What can go wrong and what are the most likely flaws in the process?

Though I've done some web/journal seaching I can't really find anything that is either a) down to my level of knowledge :oops: or b) not buried within a huge paper explaining a ton of stuff I don't need to know and sends me to sleep! :roll:

Basically I need to include it in the discussion of a feed analysis we've done in labs at Uni. So, it's a serious question folks, please don't derail and give me even more shlte to wade through that I already have.[-o<

I would PM folks directly, but in my defence, I'm unsure of who exactly to ask - just know there are a few scientificbods on the Org. who may be able to help me out.

Re: Question for the scientist/geeky types
 

From what I gather on my quick search, its an elemental analysis (so tells you the amount of hydrogen, carbon, nitrogen etc that comes from combusting the sample). Is this right?

It appears that the method only gives a rough guide as to whats in the sample, and perhaps some indication of the elvels but its not completly accurate.

I could be wrong here, I havnt heard of a "proximate analysis" before, but perhaps I have come across it in some other form...

Matt

Re: Question for the scientist/geeky types
 

:smt102 sorry.

Re: Question for the scientist/geeky types
 

Actualy K, forget that - what I was thinking of is termed "ultimate analysis", which I know as elemental analysis

Matt

Re: Question for the scientist/geeky types
 

A proximate analysis is where you just guess it.

Hth ;)

Re: Question for the scientist/geeky types
 

AFAIK it's a method of analysing a chemical by way of its properties - ie. burn it, measure its moisture, smell it ;) This is in comparisson to an ultimate analysis which is based on physical makeup where you can describe it chemically. Chemicals can have attributes of other chemicals whilst differing greatly in others so there's loads of assumptions at all levels.

Re: Question for the scientist/geeky types
 

Kinda yeah.;)

To be honest the best info I have come up with was a simple dictionary definition of the word 'Proximate':roll:

(Which is of course different to 'Proximo' - which was the last role played by the great Oliver Reed in Gladiator. :takeabow:)

Yeah - we were measuring the contents of some horse feed for values of Protien, Fat, Moisture and Ash (though there were also to other values that can be gained by this method but we didn't look into them).

Told you this Nutrition lark was more involving than 'Horses just eat grass'! :roll:

Thanks for the info though guys - it has given me another route to look down (Ultimate Analysis) so I could perhaps have something for comparison. Cheers.


SV650.org - the place for bikes, chemistry and IT advice.:cool:

Re: Question for the scientist/geeky types
 

I gues the limit is that it gives you some information, but not as much as ultimate analysis. You might want to think about other things like limits of detection of the instrument you are using, and any measurement uncertainty associated with the results (i.e. you recorded 30 units on the machine, which has a +/- figure associated with it because nothing is perfect :D)

hope this is of some use aswell

Matt

Re: Question for the scientist/geeky types
 

And you know where to come when you have that uncomfortable itch too :rolleyes:

Re: Question for the scientist/geeky types
 

Having worked in a lab that did that I should know, in theory lol. It was a while ago though. Proximate analysis is approximate, it will rarely add up to 100% but should fall within an accepted percentage for the sample matrix/analyses. There are machines that'll do the lot for you or by using e.g. Kjeldahl or combustion using the Dumas principle for nitrogen multiplied by the appropriate factor for that sample matrix type, along with Ash, Moisture & Fat (analysis must be tailored to the sample type e.g. infra-red moisture analysis for samples containing humectants, total fat or free fat, etc incl sample amount used). Aswell as Method variability, -in any UKAS accredited lab the Coefficient of Variance must be calculated for the given quality control on each run (sometimes at the start & end of the run) & it's result must adhere to statistical probabilities & recognized statistical fail patterns; there is also the method probability.

Not sure if it this does help lol