[labnetwork] Nitrogen Gas Flow Meters

[John Shott]

Pat et al:

I was asking many of the same questions on this list nearly a year ago 
and got some useful information from a variety of places including 
Michigan, Berkeley, Draper, Cornell, some of your colleagues at 
Princeton and probably a few others. A search of this mailing list 
archive from about August 2011 should show those messages.

With that information, we've selected flow meters, have them installed, 
and are coming up the learning curve in terms of collecting data, 
reliability, etc.

The first important question is whether you have a stainless piping 
system or a copper piping system or some of both.  In our case, we have 
both .... but 90-95% of our usage is on the "house" nitrogen system 
piped in copper and a significantly lesser amount is flowing in our 
"Ultra High Purity" system piped in stainless (we only have one LN2 tank 
... but at least have some additional purifiers that make the UHP stuff 
somewhat better than the house nitrogen.

However, meters clean enough to go on a copper system are going to be 
significantly cheaper than their stainless steel counterparts.  Some of 
the other labs will likely have more relevant information if you have a 
stainless steel system.

For our system house nitrogen system that is piped in copper, we have 
ended up with two different style meters:

We have "vampire style" meters that clamp on a section of type L 3/4" 
copper pipe that are built by CDI meters (www.cdimeters.com).  They are 
easily the most cost effective meters that I was able to find, seem to 
work well and have large useful flow range.  In fact, the maximum flow 
that these things can read is HUGE .... like 150 CFM.  We've used those 
primarily in our main facility where we do see some fairly high flows: 
when our tube cleaner is running, that flow meter will hit 60 CFM and 
we've got others that run 10-15 CFM day and night. (Aside: I apologize 
to those of you who are offended by my use of CFM ... I can think in 
l/min, but I've never learned to think in m3/time.)  Probably the only 
downside of these flow meters is that  you need to drill and clamp them 
onto a length of pipe that has 20-30 diameters of straight, unobstructed 
flow upstream and something like 6-10 diameters downstream.  I was 
worried about drilling into a actively used section of pipe, so we 
pre-drilled sections of pipe about 30" long, deburred and cleaned them, 
attached the meters and then cut out an appropriate length of pipe, 
cleaned an deburred those ends and then used crimp-on fittings to attach 
the new section of metered pipe.  We have a total of 24 of those meters 
installed and they have on-board totalizers.  They also have a MODBUS 
interface that we are able to poll over the network using NET-485 MB 
interface adapters from Grid Connect (www.gridconnect.com).

The second style flow meter (I think that we have 14 of these) is 
primarily used in the private labs (that we are now able to charge for 
nitrogen use) but whose use is typically smaller.  In those applications 
we selected an Aalborg XFM flow meter with a range of 0-100 l/min.  
These lower capacity flow meters seem pretty good down to comparatively 
low flows.  How accurate are they down below 5% of full flow?  I don't 
know ... but haven't been asked.  They are digital flow meters, have an 
onboard totalizer, and communicate using RS-485 (note: while most MODBUS 
runs over RS-485, RS-485 != MODBUS).  We have found that we can also 
communicate with these over the network, with a slightly different 
adapter from Grid Connect: their NET-485 device.

Note: one NET-485 MB device can be hooked up to as many as 32 CDI meters 
and a single NET-485 could be connected to as many as 32 of the Aalborg 
meters.  Depending on things like fire walls (physical barriers, not 
software firewalls ...) we actually have a total of 8 of these network 
devices to talk to our 38 flow meters.

For ball park pricing, I'd use:

CDI 5200 meter including serial communcations option: $550 each.
Aalborg XFM meter (with RS-485 built in): $1100 each.
NET-485 MB: $150 each.
NET-455: $100 each.

The big cost unknown, I believe is software to communicate with this 
stuff.  We're doing our own .... so it's free, correct? But not yet done ...

There are companies that will sell you turnkey systems, of course, but 
that is likely comparatively pricey.  Others may be able to comment on 
that approach.

I've rambled on enough, but hope that this will get the discussion started.

Thanks,

John

On 6/27/2012 1:14 PM, George P. Watson wrote:
>
> Dear Labnetwork members,
>
> Our engineering school Nitrogen cost has risen dramatically and our 
> method of dividing the cost (mostly guessing everyone's usage) has 
> become outdated.  We would like to meter each lab and facility to more 
> fairly assign the cost.  Has anyone had experience installing and 
> using these meters?
>
> Can you recommend a meter capable of maintaining the N2 purity?  How 
> much does such a meter cost, and how much is installation and hookup 
> for remote readout?
>
> What happens if there is a wide range of N2 flow rate, both from lab 
> to lab, and at various times within one lab?  Will one meter meet our 
> needs?
>
> How do you monitor the N2 use?  Do you read flow totals from the units 
> themselves, or are they connected to a network?
>
> Thanks in advance,
>
> Pat
>
> George Patrick Watson (Pat), Ph.D.
>
> Director of Operations
>
> PRISM Micro/Nano Fabrication Laboratory
>
> E-Quad J301A
>
> Princeton University
>
> Princeton, NJ 08544
>
> gwatson at princeton.edu <mailto:gwatson at princeton.edu>
>
> (609) 258 4626
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mtl.mit.edu/pipermail/labnetwork/attachments/20120627/bd4b5186/attachment.html>