The US Navy Contracting Approach Fails the Stability Test
In the 1990’s, my father was asked to attend a side shell damage on a US Navy vessel. The damage was caused by assist tug hard contact. He surveyed the damage and came back to the office to write it up. His partner Harry Ottaway saw my Dad’s commercial damage repair estimate, and said: “Henk it is a Navy ship. Double it, and double it again”. He did and came pretty close to the Navy’s claimed repair cost.
This was not a complex damage, and in actual repair process was no different than the repair of a commercial vessel side shell damage, but still it somehow cost four times as much as it would cost in a commercial setting. To my father this was a surprise because when he worked for US underwriters in Rotterdam in the 1960’s, he remembered attending to a US Navy destroyer that had a contact damage on the transom a few feet below the waterline.
He suggested to the XO that they arrange for drydocking, but the XO said: why don’t we do it ourselves with a cofferdam. The cofferdam was faster and less expensive, but also technically much more aggressive. My father was sufficiently impressed to tell me about it a number of times over the passing years. In retrospect it made sense that wartime hardened US Navy personnel knew how to deal with an issue like that.
However, that experience base had long since faded away and by the 1990’s we ran with the double and double again multiplier.
Recently we were asked to attend on another tug assist sideshell damage on a US Navy hull. The US Navy had helpfully supplied their cost estimate in the form of manhours of the various trades. Noting that it was a simple steel side shell damage that only required the removal of some crew racks and ducting, it was disturbing to note that in this estimate the cost to repair was double, then double, and then, once more, double the commercial cost to make the repairs.
Because the manhour estimate was provided, and there was a clear and simple understanding of the amount of steel that needed to be repaired, it was rather easy to calculate the amount of welding that a welder was supposed to perform in this estimate. It turned out that the US Navy estimated that a welder would produce about 1 foot of 3/16” side shell butt or seam weld per day!
It is difficult to precisely predict how much welding a welder can do in a day on a repair job, but one foot of stick welding in an hour should be easy to achieve, and with proper focus on production, much higher speeds can be achieved. In other words, the US Navy assumes that weld repairs take at least 8 times longer on Navy ships than commercial vessels.
There are associated hourly costs with welding work such as fire watches, QA and supervision that also drive the cost for this simple repair through the roof. The estimate appeared to indicate that instead of one fire watch, the Navy needs four fire watches during hot work.
When I discussed this with the Navy, the estimator noted that this number of fire watches is the minimum required for an 8 by 8 foot side shell plate repair in a single compartment.
He explained to me that this was not some commercial vessel and that a Navy vessel needed this many fire watches. Noting that the US Navy is quite successful at setting ships on fire during repair work this may be true, but maybe compared to commercial work they are putting the fire watches on the wrong jobs.
The estimator insisted that his estimate was based on prior experience and followed the NAVSEA repair estimating manual and was therefore proper and reliable.
That comment will take a bit of work to unpack. Prior experience is only valuable if it provides realistic data. One foot per day welding may have become “prior experience” in Navy repairs, but that does not mean it is realistic. To assume that prior experience is necessarily realistic is inherently flawed. If something is done incorrectly in the past, to rely on it will simply result in incorrect processes in the future.
Next one will have to consider how the Navy ended up with such “prior experience” data. In retrospect it makes perfect sense, once a contractor knows that the Navy relies on prior experience in assessing the fair and reasonable cost of repairs, they will work extra hard at increasing the “prior experience” cost. It is a vicious cycle.
Bizarrely, it is easy to break. Send a Navy estimator to a productive shipyard with a box of donuts, walk into the office and ask the manager to be able to watch a welder do some side shell welding. Measure how much welding gets done and that becomes the “prior experience”.
I suppose government contractors will then have to figure out other ways to gild the lily.
Generally, shipbuilder lily gilding is achieved through customer naivety or ignorance, but sometimes it is self-inflicted by the customer and ends up frustrating the shipbuilder.
The US Navy reportedly spent billions on modifying the Italian Constellation Frigate design to an extent that it is unbuildable and that project has now been cancelled. That is not the fault of the builder, in this case the builder is suffering from the customer’s lack of discipline.
Regardless, it all amounts to incredibly inefficient Navy construction and repair, and in the end we all pay for it. Defense is a cost we all need to carry, but inefficient defense is a recipe for disaster.
For every column I write, New Wave Media makes a small contribution to an organization of my choice. For the foreseeable future I am selecting SL7Expo. An industry wide effort to develop a Smithsonian level exhibit center for commercial maritime.
