Holdfast Training Services Trial Blog

I have now started my design attachment in the USACE’s Baltimore District Headquarters and am currently working on two projects. One of which is a ‘server room’ cooling survey at Fort McNair in Washington DC, which this blog will focus on outlining.

Fort McNair, in the heart of Washington DC. As you can imagine the traffic was delightful.

I have realised my utility, I am cheap, and so have used this to secure some responsibility early. The project budget is $50,000 and, in the mechanical section, the average engineer’s time is billed at $135 per hour. The basic sum on this gives an engineer 370 hours, but add in project manager time, vehicles and other overheads and it can soon be eroded. This is something I will research into for a further blog but the upshot is that this is now my project, and a handsome little mess it looks like to. The scope of the work currently is to write a report on the cooling within a number of server and communication rooms within the Military District of Washington (MDW) office buildings. There are 15 different rooms spread across 8 buildings, built circa 1900, all with vastly different loads and in different conditions.

After meeting with the USACE project manager, the client and the HVAC engineer for the buildings, Don Ruhl (my partner from the mechanical section) and I toured a number of the rooms. As we travelled around it became apparent that not only did the client not know what they wanted; they also didn’t know what they had in the rooms. The photos give an idea of a couple of the rooms and the varying conditions.

Don inspecting the many unsealed penetrations in a small converted basement broom closet. This room had about 12U of switches and had a retrofitted ductless cooling system. It also had a condensate drain to a sump, thus allowing the condensate to re-evaporate and continually cycle through the cooling system.

Another basement room with abandoned hardwiring to the Pentagon. About 45U of high grade servers in here. The pipe on the floor is for chilled water with heavily damaged insulation allowing condensate to form on it in the summer.

Server rooms are ideally internal within a building in order to avoid solar heat gains and also because servers don’t need a window to stare forlornly out of. They are usually sealed from infiltration; tidy, to control airflow and have some form of HVAC. The current standard for low and medium density data centres and server rooms is to use a hot aisle, cold aisle system as illustrated below. Cold air is fed from low down in the ‘cold aisle’, either through the floor or by retrofitted ducts; the server blade draws it in through the front and rejects warm air through the rear into the ‘hot aisle’. This rises and is collected by the return air system. As the photos above indicate this was not the case.

Hot aisle, cold aisle process diagram.

So what. Well given the conditions in many of the rooms, even doing a complete survey would be incredibly costly on time. After this initial assessment we need to engage in some expectation management in what we will be able to provide and re-write the scope of our work, which is currently pretty open ended. I see this largely as focusing effort on the more important rooms in terms of upgrades and identifying the risks of each room to the client so that they can make an informed decision.

In an age of BIM, I have been told that the best we can get is a floor plan for some of the buildings so it appears that even a set of out of date as built drawings are a wish too far. Due to this if construction work is ever completed on this then the contract will almost certainly have to be a design and build as the potential for change orders on a traditional contract would be immense!

Finally, a little leadership challenge.

I mentioned earlier that I had been given the lead on this project but I am working with another engineer. Don has worked for USACE design section for at least 30 years and is probably the most intelligent person I have met out here. He has no aspirations of leadership and is very happy to let me control things; however his ability to take a tangent and dive too far into the details too early are something to behold. Certainly a different management challenge from both soldiers and contractors!

Part of the role of USACE, as the client’s representative, is to conduct design reviews for design-bid-build contracts (read traditional contracts). These are done at 35, 65 and 95% with comments provided back to the project manager and design team, be it in house or a consultant, through an online system (Dr Checks). The designs are reviewed by us at construction division as well as the design division and are passed out to the clients and facilities managers, probably amongst others.

A couple of weeks ago, due to staff being on leave, I was given one of these to look at the pumps, seemingly alone. Having no idea what to do I browsed the drawings to work out what the issues might be. The project is a new 750,000 U.S. gallon water tower for domestic water and as a fire supply so my pavlovian response was Bernoullis!

After checking the answer it seemed about right although there were no accompanying design calculations to the contract and drawings so I chalked up my first comment. The rest of the checking passed with less excitement. There were a few clauses that had been missed from the contract, some ill thought out processes and demolition elements missing from the drawings. It seems a common theme though that construction division give the most comments, usually about build ability and, as discussed in the past, what is actually existing at the site.

So what have I learnt:

Hopefully I’ve done Bernoulli’s right; simplify the problem and sketches work.

Designers, it appears, live in a magic construction dreamland and it is always the same build ability issues that are picked up. By using traditional contracts USACE does assume a lot off risk and pays handsomely for the privilege if elements aren’t caught by the construction team prior to tendering. Having recently moved into dreamland, albeit part time at present, the fine detail is easy to forget.

And as ever, time spent on recce…

I am not sure whether the client usually gets to see the value of each of the subcontracts but it appears that we do. So for my FtIG project I can see that ‘the bus’ is costing us approximately $110,000 or 27.5% of the contract value. From what I have seen so far this pays for two slippery salesmen and an OAP in a portacabin who is going to run the site. Although we are supposed to get the information it has been accidentally added into the emergency contact submittal, which was clearly part of a larger spreadsheet, but you can get the idea

Contract values

Flippant as I may be this clearly also considers the various overheads that have been incurred not least the cost of tendering for the project and the 4 months of paperwork that has been done prior to getting out onto site. This in cash-flow terms is a significant burden for a small company. Also in there will be the risk for things not going well, something that is quite likely in replacing an old system. I am not sure whether Greg or Steve have any thoughts on the general mark up on something like this.

As far as progress goes, yesterday I met the aforementioned OAP who is to be the superintendent, Quality Control Manager and Site Safety and Health Officer (SSHO) and some materials were delivered. I took the opportunity to get some shots of the interior:

Left: Air conditioning unit not being replaced. Right: Two boilers being replaced.

Left: Three pumps, to be replaced with two pumps.
Right: Domestic Hot Water Heaters to be replaced.

Safe in the knowledge that everyone will be too busy fawning after Mike’s baby…

I just reviewed the FtIG boiler project schedule, again. Now, on version 4, the contractor has lowered my expectations sufficiently that I am now focused only on the money. Their timings for installing pipe work are unrealistically short and everything appears to be on the critical path. These, though annoying, are their problem as the contractor is responsible to ensure there is no gap in service for hot water, heating or cooling within certain seasons.

So resigned to failure on the time side I have turned to look at the money. USACE pay for definable elements, rather than the alternative of paying for percentage complete on the whole job. From the pay reviews that I have conducted for the Waste Water Treatment Plant this appears to be a better form of holding the contractor to account for finishing up individual elements. By paying 98% of an element that isn’t quite finished it leaves a virtual snagging list in dollars. These are more likely to be completed by the next pay period rather than being forgotten until the end of the project, when especially if there is a staff turnover, they get swept under the rug. We will pay for stored materials but, it has to be a significant element, the contractor has to show ownership and we prefer if it at least in place. For this project the boilers and water heaters may come into this but certainly nothing else. Again, for the waste water treatment plant we have paid for stored materials for a $60k generator and a $130k pump set so that really sets the magnitude. I’d be interested to hear what everybody else’s clients’ are happy to pay for on that?

So what’s my issue with the FtIG values? I have two:

To me definable elements are like a product breakdown structure (PBS) and when it comes to pay time I want to be able to go to the element, look at it, tick a box and pay the contractor. This is not me being an idle box ticker rather ensuring that each element is well defined, and I assume this will help them with their sub-contractors. The FtIG programme has not been built from a PBS so I can foresee the arguments coming at pay time. The lesson from this is do a detailed PBS, even for a simple job, because it has ongoing utility beyond just being the start of a programme.

I remember Steve Payne hammering the PBS into us in phase 1 but this experience and if I had any doubt then it has gone. The issue of everything being definable is important too in order to communicate the plan to different people. I am, reasonably, sure that the sub contractor knows how to do this work, however they have not managed to communicate this effectively to the contractor who, in turn, has not managed to communicate it to me. I am sure this is a lot worse in the larger projects with more people involved.

Front loading. The contractor admitted before that the project was front loaded, we all laughed and I told him that we wouldn’t accept front loading. They went away and when version 2 came back it was worse rather than better. I am aware that the contractor requires cash flow. Working for the Federal Government cash flow is not a problem, however I need to ensure that if the contractor goes bankrupt that I still have the money left in the project budget to put the remainder of the contract back out for tender. The solution to us meeting a better agreement actually came from the contractor letting me loose on his values. I can’t pay them any less than the contract value, all I am doing is moving around what the money is tied to. I am sure there will still be more iterations required before we come to final agreement but it’s a start.

In terms of submittals, over here the Navy ask for a ‘schedule of values’ as well as a schedule (programme), but USACE ask for both as part of the same submittal: the schedule. I’d be interested to hear how others have seen this being asked for by clients? For checking on the client side I think breaking them out would be best.

Apologies for the lack of pictures, I have avoided putting the programme in as it is little changed from a couple of blogs ago, and looks pretty difficult to see on a computer screen.