Design First, Build Later

Design First!

Sounds simple. It makes sense. But far too many people do not do this. They put together an idea, a process, but fail to take the design to completion before starting to build. Cutting corners always comes at a price. There is an old saying in Project Management, “You can have it good, fast, or cheap. Pick any TWO”. Cutting corners will bump your cost, decrease quality, or possibly both.

A while ago I was tasked with designing and building a system for groundwater and soil remediation. The science was basic. The groundwater contained multiple hydrocarbons including both light ends and DNAPLs and the soil above the water line also contained volatile hydrocarbons that were creeping up from the contaminated water. We had made many of these in the past, and since. The issue was not the remediation science; the issues were in the complexity of the system and the physical constraints of the project.

I have made countless containerized systems, where you fit the equipment into standard, or in some cases high cube, 40-foot-long ISO shipping containers. This is a very effective method as it allows the system to be easily shipped anywhere while providing a safe, robust enclosure for the system to operate. ISO containers are also quite plentiful, and reasonably priced, making the economics of their use is advantageous.

The First Issue

The container needed to be compartmentalized. We were going to be bringing in groundwater that contained various hydrocarbons, some of which were light end and volatile. The part of the container that housed the incoming feed piping from the wells, the oil water separator, and the double wall tank for hydrocarbon storage was a Class 1 Div 1 enclosure. The remainder was general purpose and housed the valve banks, the controls, and an air compressor to run the sparge lines.

Noise was not allowed. Seriously, the system was to be used close to a residential area and we had to make it run at below 60 dB. This complicated the space issue. We built a small room around the compressor, and then had to install multiple acoustic attenuators within the container including a Helmholz Resonator on the ventilation fan and on the intake louver for ventilation. Each box was six feet long, four feet high, and two feet wide.

Real estate was at an absolute premium. Which leads to:

The Second Issue

The second issue tying in with the first was the complexity of the system. We needed to control the whole system from within the General Purpose room. Therefore, every incoming line had to pass through the barrier between the two rooms, receive its necessary process controls and be measured, then pass back to the Class 1 Div 1 room. We accomplished this with two separate valve banks. One on either side of a quarter inch A36 steel plate with full couplings welded into it. The plate was a elegant solution that was removable for maintenance.

The first picture shows the lass 1 Div 1 side of the enclosure. The operators needed to be able to monitor pressures in real time and have full, instant shutoff control of each line from within that enclosure. The solution was simple, robust, effective, and practical. It looks simple, and it is, but it is what we had on the other side that makes it somewhat insane.

And here we have the backside. Each line incorporated both controls and instrumentation. The operators have full system knowledge and oversight from the control room. Being general purpose, the solenoid valves were demonstrably more economical than their Class 1 Div 1 counterparts would have been. Considering the number of them, this saved the client a small fortune. The blue plate in the first picture can be seen high up on the wall. This was an extremely complex piping spool to build. The secret to the success of the build was the effort put into the design. Design first, then build.

And just to be complete, here is the third valve bank in that system.

The SolidWorks Advantage

The shear number of components, the routing, and the supports make this virtually impossible to design by hand. Not only did everything have to fit; it must be easy to operate and easy to maintain. We always design with the operator and maintainers in mind.

Using the 3D CAD package SolidWorks, we were able to complete a full design and test that design for operational requirements. It is also quite handy that SolidWorks will generate an accurate and complete bill of materials to make Purchasing happy.  There are a number of high-quality 3D packages out there, I simply prefer SolidWorks.

Moral Of The Story

Generally, the later in the project lifecycle that we make mistakes, the more costly those mistakes are. For example, any work done in the field costs at least three times as much, in both time and money, as work in the shop. It is better to make your mistakes on a computer screen, where the correction only costs you a few minutes, than it is during testing when you have to rip out a valve bank and delay delivery to the client.

Design first, build later. That sounds simple, but many do it backwards, and pay for it. If halfway through the build your tradesmen are coming to you with suggestions of how to better design the system, you have already failed. A good designer will always run his ideas by his colleagues on the floor and get their input while the design is still on the screen. These guys work hands-on eight hours a day, five days a week. They know what works and what doesn’t. They are one of your most valuable assets as a design engineer. Once completed, hand the drawings to the manufacturing manager to be red-lined, then to the quality manager for a similar review. Yes, it will take an hour or two of their time, but they will invariably catch mistakes, or make suggestions, that will save many more hours and dollars during the build, and may save frustrations for the operators down the road.