[Gasification] combustion gas analyzers, advice needed

[jim mason]

On 2/27/07, doug.williams <Doug.Williams at orcon.net.nz> wrote:
> Hi Jim, And Colleagues,
>
> If I read your need correctly, you want to analyse the gas you make, rather
> than what is in the exhaust gas after you have burnt it.  The combustion gas
> analysers, are not suitable, and selective gas analysers expensive. (ours
> cost US40,000).
> The real problem with these systems, is that they contaminate the sensor
> probes very easily, and  need recalibrating, always when things get
> exciting!!! (:-)
>


well, i missed that one.  you are right.  these units are for
combusted gas, not precombusted HC fuels.  the sensors would likely
plug terribly if measuring 10x percentages, and not trace ppm.

i assume it would be the same for a smog check type smog machine.
they are really combustion analysers like the above.

but then again, these analysers should work for co2, nox and o2.
these levels will be lower or about the same as post combustion, so
the sensor range should be ok.  really all you need to measure to
assess reduction bed health is co2, as you note.  but without a h2/co
measure, you aren't going to see the h2 rise as temp drops.  of course
we can also just directly measure the bed temp with thermocouples
(which is another message asking for advice on that too)

your diy method below is very interesting.  i'll go look it up in more
detail.  i remember their being some other options for doing some of
these in the books.  but i forget them at the moment.  i'll go look at
those again soon.

but in the meantime, if there are other favorite methods to consider
for raw gas analysis by others, i'd love to hear about them.

j







> What I offer below applies to high temperature packed bed gasifiers using
> air as the oxidizer.
>
> In the early days of our research, we used a German system, that used a
> liquid filled flask, to measure CO2 in flu gas. You just pumped about 10
> squeezes though it, and turned it upside down. The CO2 then gave a read out
> on the liquid level. I believe they are quite cheap now. While this is not
> laboratory measuring, it provides you with a good idea as to how the
> reduction zone is performing, with 7-16% CO2 within a ballpark range of
> acceptable CO2. As Nitrogen dilutes around 51%, the rest will be combustible
> gas.
>
> If you then take this gas and cool it right down without condensing visible
> tar, this means the CH4 is  below 3%, and then you have mainly H2 and CO.
> Then to guess how much H2 is in this balance, test lighting the gas with a
> spark from a flint gun, (mixed with air of course). If you get instant
> ignition, a good guess is that the H2 will be around 15-18%. The
> flammability is always a good indicator of the H2 level. In an engine, this
> translates to easy starting, and fast response to load change.Gas of this
> quality has around 135-140 btu / Nft3.
>
> All good quality producer gas (tar free) has a similar gas analysis,
> although it is possible to tweak things using mysterious black boxes, and
> cylinders that clank in the night ( I'm not kidding). For most of us mere
> mortals though, all the above applies, and if used intelligently, provides
> enough knowledge to get to the next stage of our interest. However, if you
> are going to get serious and  stating gas specifications, then you need to
> send samples for approved lab testing. Darren Schmidt can offer you these
> facilities over in Grand Forks, N.D.  dschmidt at undeerc.org
>
> While many are obsessed with gas analysis, it hides a multitude of sins if
> taken at face value, and alone should not be taken to a prove a gas making
> process. It only records gas quality at one moment in time, at an output in
> that moment, no more.  Many factors affect the analysis, and that is why we
> see such big variations in the literature. The ultimate gas test, is an
> engine doing work, and if you can use it without killing it, you are on your
> way.
>
> Hope this helps.
> Doug Williams,
> Fluidyne Gasification.
>
>