Payday loans
Home People Gerry Spain Setting up a new site
Setting up a new site PDF Print E-mail
Written by Gerry Spain   
Friday, 25 March 2011 12:51

Some time ago, I was asked for an opinion on key requirements and issues related to setting up a new site for atmospheric monitoring - this is what I came up with.

 

Setting up a new site for atmospheric measurements (primarily trace gases)

Most sites seem to develop organically; in other words development is mainly in response to immediate needs without any definate plan or design. The initial physical infrastructure can determine how a site develops and will almost certainly be the cause of limitations which only become apparent long after the option to knock down and rebuild is gone. Spending some time on even a rough plan can avoid making some decisions which, in retrospect, are really dumb. So how do you come up with a plan when you have no idea what you are planning for? Well, start with your own measurements. No matter what you're measuring, you will have some specific requirements like the inlet needs to be 10m above ground or it needs to be straight or you need uninterrupted exposure to the ocean or the sky etc. Secondly, what ancillary measurements will you make and what's required for them? Thirdly, will you have collaborators running other measurements - what do they require? Finally, what about possible future measurement requirements? The latter is difficult to estimate but it will depend a lot on what the ultimate goal for the site, if there is one. If, for example, it's a potential GAW site, then there are a range of measurements which GAW require/recommend depending on the type of site and they are outlined in various GAW publications eg a useful overview can be found here - ftp://ftp.wmo.int/Documents/PublicWeb/arep/gaw/gaw143.pdf
The aim is not to make detailed plans for measurements you might make in the future, just to be cognicent of potential future needs and avoid making decisions which amount to shooting yourself in the foot.

Visit other sites to look for ideas. Concentrate as much on what is bad about a site as what is good. Ask the operators about what they would change if they could, what mistakes were made as well as how they get around them. Equally, ask about the best features of the site, what should definitely be included in your new site. No matter how bad the site, restrain yourself and make some positive comments - you'll learn more.

When it comes to the site layout, start with the groundwork. Aim to do as much groundwork as is reasonable to allow for future expansion. Assume there will be expansion and plan the site layout with potential future measurements in mind.
If it is intended to have more than one building/container, ensure a suitably sized area is cleared and leveled from the start. It is much better to keep things on the same level if possible.
Allow for a staging/unloading area, preferably on the same level as the buildings. This should be leveled with concrete or tarmac - gravel is useless for moving things with pallet trucks and trolleys and should only be used for parking/turning areas
For rainy sites, allow for drainage and try to ensure buildings are not likely to be flooded eg don't site them where water is likely to flow into or collect or use french drains to divert water.
Allow for plenty of connectivity between labs - install separated ducting for shared services eg telephone, network, power, water etc. Ensure the ducting is as straight as possible and is big enough to allow ease of snaking cables/pipes through and space for expansion
Think about the positioning of any towers and associated guy wires and anchor points. A taller tower is easier to deal with - the guying points on the tower are higher and the anchor points are further away so there is more 3-D space to play with near the tower.
If a walk-up tower is not required, a self-supporting unit which is hinged so it can be lowered for servicing might be more suitable. Obviously there needs to be a clear area for the tower to lower into.
Think about positioning of structures with respect to measurements and include potential future measurements. eg if solar measurements are to be made, ensure there is a location available where they will not be shadowed or if a backup generator is to be installed, use local met to determine the best downwind position.

Next the building itself. In general the options will be a container lab or a solid building, each with pros and cons.
A container lab can be completely built off-site and delivered more or less ready to go. It's flexible in that it can be relocated relatively easily and can be "extended" by adding another container on top or to the side. Downsides are limited space, rust issues, termites etc in wooden sections
A solid building is suitable for a long-term site, is more durable (well, concrete is...), can be built to fit needs, can be designed to be extendable and provides some local employment. Downsides are mistakes in siting or design are less easy and more expensive to rectify.
In either case, the lab will need to be designed more like an industrial installation than a home or an office - it has to be practical rather than pretty but it should still be neat and logically laid out.

Some specifics;
Container lab
- Use an aluminium container, definitely not steel. The extra initial cost should be balanced against reduced maintenance and no need to replace it after a few years
- Stick with the standard 8x20ft size (or 8x40ft). It fits on standard trailer and takes one container berth on ships (a 9x21ft will occupy 4 berths ie 4 times shipping cost)
- Clad insulation on outside to create some extra space inside. The insulation needs to be designed such that it can be fit after deployment and in small enough sections to ship in the container
- Ensure the container is leak tight before installing any equipment and double check weak points like rivets.

Solid lab
- If there are 2 or more floors, install concrete or reinforced floors for extra load bearing
- Consider a lifting system to shift heavy equipment to upper floors
- Install some ducting through floors so you can retrofit cabling/tubing etc

Other considerations;
- Daylight is nice but windows are a security issue and another place to leak in a container lab. Investigate light pipes or skylights as a lab without any daylight is horrible to work in. The positioning of any windows is important so instruments or temperature sensors do not get direct sunlight and  solar gain is minimised (or maximised if you're in the arctic...). Tinted windows may be an option to consider.
- Design with local power and local standards in mind as it's easier to source/substitute local replacements if necessary. Applies to all fittings and fixtures (eg a door handle you can buy in Home Depot in the US for a few $$ might be nothing like the usual handle used in Rwanda). If there are no local standards or if they are crap, then design for whoever will be looking after the place (ie if it's a UK lab, then follow UK standards)
- Design a simple shelter to completely shade the lab from the sun. This will reduce thermal loading and also keep heavy rain off. Obviously this applies to a hot site but a similar approach can be taken to accommodate the weather at any site. A very wet or heavily snowed site may need a large overhang or porch to maintain a dry(ish) area outside the lab
- Separate noisy pumps etc from main lab if possible. Best option might be a separate container which could also have section for storage of gas cylinders and other materials
- Include a section in lab for gas cylinders, generators etc
- Lab doors should be extra wide
- Use cable tray system or similar to route gas lines to various parts of lab, colour coded shutoff valves at delivery points, same colour code markers at various points on the lines themselves
- Power. Use either a large UPS to supply whole lab or use individual UPS per instrument/cluster. Ensure that diagnostic info is logged from each UPS. Install power analyser at the supply to lab and log diagnostics from it too. Lots of issues are caused by power fluctuations and it's easier if you can definitely pinpoint these.
- Flooring. Lab or hospital grade lino with upturn to ~6in. ie no joint at edges where gaps can develop allowing bugs in or small ferrules to escape
- Island benching is useful for access to rear of instruments but uses up space
- Floor to ceiling 19in rack would be useful for certain instruments. If the instruments can be moved, use a rack with lockable wheels to allow access to the rear 
- If installing a system with many components, look at reducing the space requirements. eg with the Medusa system, a lot of it is external equipment which might be made smaller (eg trap supply) or may not be necessary (eg can the MS roughing pump also back the diff pump, maybe with a slightly bigger pump?). Design external components to stack neatly or fit with each other.
- Create some space for your tools and spares. Stackable boxes are good. Also dedicate a bit of bench space to work on repairs etc.
- Air conditioning.  Connections in A/C units will almost invariably leak, specify a hermetically sealed unit with all joints brazed. Check that the unit can be configured to maintain the temperature to within the limits you need. Modern inverter units typically are set to a differential of +/- 2deg which can often be reduced to +/- 1deg. Ensure the A/C does not blow directly onto instruments. Try not to have stuff under the A/C in case the drain blocks and it leaks all over the place. Log the lab temperature at least as frequently as your most frequent measurement.
- Tubing and cabling should all have a drip loop outside where they enter the lab (obviously upwards, not necessarily through the floor)
- If aerosol measurements are to be made at the site, suggest a separate lab. Aerosol measurements have different operational requirements and many of the counters use butanol as condensing fluid so you don't really want lots of that vapour floating around in a gas measurement lab
- Passive measurements (radiometers and the like) don't require a lot of lab space. Typically enough for a computer and a few spares is all that is needed, (the business end is always outdoors) so can be fit in anywhere but if the site is to include a wide range of measurements, then a separate space for such measurements would be useful & could double as office space.
- Noise (again) - suggest putting in a little effort to minimise it. Ears are important diagnostic tools but of little use if you can only hear noisy fans and pumps.
Use silencers on pump outlets or vent them outdoors if you cannot house them in a separate space
Replace noisy fans, use low noise/lower speed versions or thermostatted fans if possible. Some blade shapes are designed to create less noise, removing protective grills sometimes makes a noticeable difference
- Assuming there is a network connection, create a private network for your own instruments/computers and isolate it from other users of the site.
Set up Skype so that people don't have to worry about the cost of phone calls. Skype on Linux computers can be a pain (maybe not much better on Windows machines) and every time you upgrade, you risk breaking it. You could dedicate an old PC to be the Skype box and never upgrade it as long as it works. Wireless headphones/microphone mean the operator is not tied to the PC while talking.  Alternatively, install a wireless access point and provide an iPod with a microphone on the headphones to be dedicated for Skype usage. It might not be cheap but the iPod is also portable so you can talk people through stuff while they're around the back of a machine. Portable makes it easier to steal and you don't have video so it's not a perfect solution. A dedicated netbook might be another option worth looking at.
- Ancillary measurements. If you need or would like other measurements (eg met) don't assume someone else will provide it. If it's on-site and you can access it, OK but otherwise install your own met station or whatever. Assuming you will run a Medusa, I'd also go for ozone, CO2, CH4, CO and maybe N2O
- Have some basic rules which all measurements adhere to eg use a common time zone, use a compatible data directory structure and file naming system, give instruments/computers somewhat meaningful names or maintain a host table with descriptions of what each computer does.
- Maintain an electronic library of useful reference material eg instrument manuals, shipping procedures, shipping documents, useful contacts, sampling logsheets. Also diagnostic information of common interest eg UPS or power analyser logs, lab temperature records etc.
- Shipping - you will need to find a reliable way of shipping spares, standards, compressed gases etc in and out of site probably using a freight company rather than a courier.

Last Updated on Friday, 25 March 2011 15:01
 
Copyright © 2017 Atmospheric Physics Research Cluster. All Rights Reserved.
Joomla! is Free Software released under the GNU/GPL License.
 
None