Kawasaki Heavy's efSET news

[bikkuri bahn]

Been awhile since this proposed trainset has last been in the news. With the PM's U.S. visit, it looks like this model will be officially offered for the California project. Apparently tender bids by the various manufacturers will occur sometime this year.

Kyodo News report via the Japan Times (yecch...):

NEW YORK – A consortium of Japanese companies is planning to propose Kawasaki Heavy Industries Ltd.’s bullet train technology for use in a high-speed rail link in California, sources with knowledge about the plan said.

Prime Minister Shinzo Abe is expected to make a pitch for the technology on an official visit to the United States that begins Sunday, ahead of bidding likely to be held later this year, according to the sources.

The technology is based on Kawasaki’s efSET, short for environmentally-friendly super express train, and is designed to achieve an operational speed of 350 kph.

 

http://www.japantimes.co.jp/news/2015/04/25/business/japanese-consortium-to-propose-kawasaki-heavy-bullet-trains-for-high-speed-line-in-california/

 

...efSET, something like a "super" E7 set?

Edited April 27, 2015 by bikkuri bahn

[Jace]

The article mentions that there is concern about bidding on price against Chinese firms. A look through the draft RFP documentation (http://www.hsr.ca.gov/Programs/trainsets/index.html) shows however that this contract will not be awarded to the lowest price bidder, a cash flow model is used instead. The proposer with the lowest life-cycle costs (capital, operating, maintenance) gets the highest financial score. This is then combined with a technical rating to come up with the overall score. This to me helps out companies like Kawasaki and California based Siemens.

 

The use of an NPV model however tends to invite delays as a contractor is encouraged to have big payments/costs occur as late as possible. This is particularly the case if your internal cost of capital is much lower than the one called out in the model. CHSRA calls for delivery and acceptance of the first 16 trainsets within 4 years - a very aggressive schedule considering that the trains, and most of the parts, need to be made in the US. Delays are penalized at up to $38,000 per trainset per day but overall delays are limited to 10% of the contract value. The tight schedule will force companies into including at least some of these delivery penalties in their price. Taking this further, I can build all of those penalties into my trainset price with a plan to deliver three years late, I'd break even on capital (compared to no penalties added) but come out ahead in the model as all my O&M costs are now pushed out three years. If I dare, I could plan on delivering four or even five years late. Penalties would be capped out and my NPV would only go lower. I'd never show this to CHSRA but what they would see would have the same bottom line NPV. The biggest risk for me would be added politcal pressure to build ahead of my internal schedule. (I wonder if this is how the whole IC4 mess got started).

 

One way to fix this is to include revenues as well as costs in the model. Two years of lost passenger revenue would surely far outweigh late delivery penalties. We'll see if this gets addressed in the final version of the RFP.

[kvp]

The operating and maintenance costs are weighted in, but not calculated into the price. This means if a trainset is late, then the operating and maintenance costs are simply not paid for the duration of the delay. This means it can not be used to lower the price of the sets, only to lower the income of the seller. Only the 10% delay penalties can be calculated into the bid, by raising the unit price of the sets.

 

Kawasaki however plans to build the first 1-2 sets in Japan and then assemble the rest in the US. The key is assemble. Only the car body jig and the stir welding equipment have to be moved to the USA and the rest of the parts (bogies, motors, equipment, even seats) shipped fully manufactured but unassembled. This means most of the train will be assembled in the US, but only the large, expensive to ship, but easy to make parts will be actually made in the US. Since this constitutes the majority of the weight of the train, it can be said that most of it will be made in the US. However most factory equipment and knowledge will stay in the japanese factories. The same trick was used for the Siemens locomotives and nobody is complaining, especially that the chinese would do the same.

 

Also 4 years is more than enough, since there is no need for new technology, just make a bunch of E/W7 parts and put them into the customised shell. The dedicated right of way means the winner don't have to add all the extra weight for crash safety either, which means minimal or no redesign is needed. On the other hand, Siemens has a US export velaro design on its shelf and a very similar (but probably more quake prone) safety and signalling system that is in use in Germany. The chinese have both the Siemens and the Kawasaki technology from earlier purchases and could accept a loss of money on the contract, just to get a larger market share. However their technology contains a bit too many shortcuts and copy errors.

[Jace]

The NPV model includes O&M costs from contract award to final disposition of all trainsets. It also includes milestone payments (i.e. capital costs) and a revenue loss factor if the trains have fewer than 450 seats. There is nothing in the model that deals with loss of revenue over any delay, that is handled strictly through the damage penalties. Late delivery will push the milestone payments out and gives you zero O&M costs for those years, both of which result in a better (lower) NPV. If you want to show a model with on time delivery, then you can capture that reduction in NPV (from the other model) by lowering the vehicle's price and/or by lowering projected O&M costs. The bottom line is that the better the NPV, the better the financial score. Vehicle price is a factor but when that price is paid is critically important: technical counts for 30% of the final score, financial 70%.

 

The Siemens locomotives had to meet a far less stringent form of Buy America than these trainsets. The locos have to be only 50% US, the trainsets will be 100% US content (standard FTA funded car projects have a 60% minimum). Components are not considered to be US if they are assembled here, they have to be manufactured to qualify. There will have to be significant technology transfer to make this happen - the whole intent of this form of the regulation. Amtrak and CHSRA petitioned the FRA for Buy America waivers on the prototypes only. Both waivers were granted (here's a link to California's: http://www.fra.dot.gov/eLib/Details/L16036). Kawasaki has built prototypes for many US projects in Japan, this is (or was) standard practice for them until the FTA changed their thoughts on Buy America. Other proposers, including Siemens, are surely going to do something similar. The cars themselves are likely to be as off-the-shelf as possible, the problem will be setting up manufacturing here for all the parts and doing so quickly so that all units are accepted (not just delivered) within four years. And note that there will also have to be quite of bit of qualification testing done before acceptance of the prototypes (this alone will probably take one year). I see this as a very tight and unrealistic schedule. I expect that this will change in the final version RFP.

[kvp]

 

Late delivery will push the milestone payments out and gives you zero O&M costs for those years, both of which result in a better (lower) NPV.

So are you saying that the maintenance costs are a fixed sum between the proposed delivery date and the end of the contract and will be paid in full regardless of actual delivery dates? Logic would dictate that the maintenance costs should be paid for each set only for their active time range. This means if the manufacturer is 2 years late and the contract was for 10 years of maintenance, then the actual amount paid will be 80%. This would mean if the trains are only delivered around the end date of the maintenance contract, then no money would be paid for maintenance. Another options is to start the contract on delivery, so if a train is delivered 2 years late, then the end date of the maintenance contract will be moved two years too for the same amount of money. This would discourage any late deliveries, since it will increase costs for the manufacturer for with no benefits.

 

Imho the 4 years from start to start of service is plausible, but only if the made in america is only chekced one layer deep. This means if a motor is assembled in the USA, but the parts were made elsewhere, then it could be accepted as US made. Same is true for parts like electronics, since there are no high power FET factories in the USA, so it's not possible to make everything (from the silicon sand level) locally. Nobody will build basic heavy industry for 16 sets, but most companies would build an assembly plant that can be moved elswhere if there are no new orders. The same is true for other projects, like the USAF stealth fighter project, which has both japanese, chinese and korean electronics, since currently there are no other countries who have any factories that could produce them. The US and the EU currently doesn't have the minimal capabilities for sustained independent industrial production and is fully dependent on international trade. Japan and South Korea has a similar problem with fuels and raw materials.

[Jace]

There is no maintance contract as part of the vehicle purchase, that will come later (presumably). The O&M costs are however a factor in the financial analysis of the vehicles. These costs are to be entered in as they occur and then discounted to get their present value. This is why the timing is important - the total maintenance costs will not change if they are delayed a few years but their present values will go down. The same holds true for the vehicle prices. This is a weakness of this type of analysis. Now that I think about it, even if you include revenues lost for late delivery, the carbuilder can pretty much ignore the impacts. These are losses to CHRSA and not the carbuilder hence really not their concern unless you tie the damage penalties to the actual revenue losses. I doubt anyone would want to sign up to a contract with penalties structured like that unless they somehow shared in revenue gains.

 

The Buy America rules require manufacture here involving US manufactured subcomponents. It's pretty easy to figure this out on a weldment or a door operator but it can get fuzzy on electronic equipment as there usually is not much more than domestic assembly. In the end, the carbuilders leave it up to their suppliers to tell them everything is ok. The content rules changed for high speed equipment as there is a strong desire to bring production here, eventually moving the engineering jobs to the US as well. These are the sorts of high paying jobs that everyone loves. How this happens with a couple of fairly small orders remains to be seen. But I can see that this desire will cause suppliers to follow the rules more strictly than normal as anything questionable can very easily become political fodder. I'd personally prefer that the focus be on buying well built and well performing trains.

[Sacto1985]

My guess: the proposal from KHI is based on the JR East E5 trainset, with a slightly different duckbill nose profile to push the top speed to 330 km/h (205 mph). Given that much of the California HSR route has to be built like its Shinkansen equivalent in terms of earthquake mitigation, why not just adopt a variant of the well-tried E5 trainset?

[katoftw]

It is more like a 800/E7/W7:- 

[Sacto1985]

Since KHI has built N700 trainsets for JR Central, I wonder will the Texas Central Railway proposal use the erSET trainset design, one that is a cross between an N700A and E7/W7 trainset?

[bikkuri bahn]

Shinzo and Gov. Moonbeam checking out the E5 simulator in a hotel ballroom:

https://www.youtube.com/watch?v=3D0a1UhZ5xE