Wednesday, November 28, 2007

Poker and the 520 Bridge (bluffing won't get it done)

In poker, smart players look for ‘tells’ in their opponents. It can be as subtle as putting down the cards front edge first when you have a good hand, to yelling “Whoopee! I have three aces!” “Tells” can let you know what is really going on.

The 520 Bridge is not going to sink or fail and the state has given us abundant tells that let us know that this is the case. Either that or they are even more stupid than I thought possible.

If the state thought there was a likely chance that the bridge might sink, they would install a set of railroad crossing arms and some warning lights at both approaches. If the bridge sank, the arms would come down and prevent anyone from continuing on to the bridge deck. But they haven’t installed them. And they won’t. In fact, if they really thought it was even possible, they would string an electrical wire the length of the bridge; if the structure broke, the wire would break and the railroad crossing arms would come down.

When it became clear that the ferries on the Port Townsend/Keystone run were seriously dangerous, the Governor stepped in and shut them down. But the 520 Bridge is business as usual because there is very little chance of it sinking.

The state isn’t even doing anything about the bridge safety. There aren’t even rubber floatation bumpers around the piers even though a tugboat hit one of the concrete posts years ago. They could fill the hollow piers with concrete, substantially improving their strength, but they do not. Even something as simple as installing more and better lights on and around the piers isn’t being done.

And if the state really wanted to do something about the 520 Bridge, they would go out and buy a bunch of concrete floats like the kind that are used for marina docks. They would put a string of these floats tied together with some room in between each one and place this floating breakwater off of the sides of the existing bridge. They could be attached to the anchor cables that go from the bridge to the huge concrete anchors on the bottom of the lake. This floating breakwater would prevent the waves from splashing up against the bridge. Currently, the floating bridges bow twenty-two inches in the biggest storms. This flexing out and then snapping back is the worst thing happening to the structures. But the state does nothing to mitigate the damage. It can be done for a lot less than $4 billion.

Which brings up another interesting question: if we could build the replacement I-90 bridge for under $200 million, how come this replacement is going to cost $4 billion?

Saturday, November 17, 2007

The Sound Transit light rail segment to the airport will never work.

This fourteen mile piece of light rail is one of the worst designed, least thoughtful and just all around horrible transit projects ever conceived. Most transit projects have a few problems here and there; this mess has a major problem every mile.

It is supposed to ‘go to the airport’. It does not even do that. The station at the airport is, in fact, over 1,000 feet from the nearest concourse door and some people will have to walk over a half mile to access the airport from where the train lets them off. The port commissioners who run the airport had promised to install a moving sidewalk, but the path goes through the parking garage and the seven foot high headroom will not allow for a moving sidewalk to be built.

The light rail was never intended to move air travelers from the city to the airport; it was always designed to move employees of the airport—mostly very low paid employees—from parking lots to the airport. A major transit hub like the SeaTac International Airport has upwards of ten thousand employees. The airport management does not want these people parking in their garages, even if they could afford it. The garages are for the airline passengers and the parking fees are a good source of revenue. But the light rail train is not for airline passengers, it is for the employees of the airport.

The airport management is so confident that almost no airline passengers would be willing to schlep their bags and themselves the 1,200 feet average walk to the train station that they are planning to build a $400 million garage for car rentals off of the airport site, spend $17 million buying a fleet of buses and spending another $17 million a year operating the garage and bus fleet to ferry car renters from the concourse to the off-site parking garage rather than ask them to walk to the train and take it to a parking garage.

The train will leave the airport station and head to downtown Seattle, twelve miles as the crow flies, to the north. But instead, it heads due east for some five miles to the southern end of the Rainier Valley. This was done to deliver service to the most economically disadvantaged parts of the city. This may have been done for admirable purposes, but it means that the train is scheduled to take fifty minutes to get from downtown to the airport and the real time will be more like sixty minutes and more. Part of this due to the fact that through most of the Valley, the train is running on the surface down the middle of the street. Nationwide, surface urban light rail moves at an average of 14 mph—one mph better than buses. Sound Transit will install fences on both sides of the track but that will not noticeably speed up the trains. Trains on the surface do not have to slow for the cars, buses and people who are literally in the way. Urban surface light rail must slow because there may be cars, buses and people in the way. If there is a car waiting at the traffic light, the operator of the light rail must slow, in case that car moves. Once the car actually moves from the intersection onto the tracks, it will probably be too late to begin slowing for the train. In fact, light rail has a terrible record for accidents. After all, trains do not stop very quickly and they cannot swerve at all.

Then, the train turns eastward and enters a tunnel under Beacon Hill. There is no rhyme or reason for this. Beacon Hill is misnamed; it more properly should have been named Beacon Ridge, for that is what it is. It is a five-to-seven mile long north-to-south ridge from near downtown Seattle to the south end of the city. It is steep, tall and sandy. It is lousy for tunnels. But that is where Sound Transit built a one mile long, three hundred feet deep tunnel. Which is a little strange since the ridge is only about a quarter mile wide. The planners actually had to make the tunnel diagonal to make it a mile long. And when it gets through the tunnel and to the other side, it arrives at the maintenance and storage yards. Here, the train turns ninety degrees and heads north into the city. Now, there is a perfectly feasible and available route around the north end of Beacon Hill which would lead the train right into the downtown. This route would necessitate the building of a spur line from downtown to the storage and maintenance yard of about three miles in length, but that is a lot less expensive than a one mile, very deep tunnel through sandy soil.

But that is not the worst of it. They saved the worst for last.

While the nearly the entire route, design and building of the fourteen mile Sound Transit light rail train is flawed, the truly horrifying part is the attempt to convert the Downtown Bus Tunnel into a bus-and-light-rail combined tunnel.

Back in 1983, the city of Seattle, with substantial help from the federal government, built a one-mile tunnel under the downtown streets to accommodate buses and eventually, light rail trains.

The premise of the tunnel was that buses, by being separated from the hustle and bustle of traffic, could come off of the freeway and go into the tunnel where five stops within the one mile underground would free the buses to move quickly and efficiently through the downtown core.

Nice theory. In practice, much was left to be desired:

--There was no way to build a simple straight ramp from the tunnel to the freeway and vice versa. At certain times of the day, from the north, the buses had five traffic lights between the off ramp and the tunnel.

--Because of the noise of the diesel engines, the buses had to ‘switch’ to overhead electric wires. Thus, the bus that went in the tunnel had two propulsion systems, they weighed more than any other vehicle on the road (over 100,000 pounds), and they had twice as many breakdowns as other buses. They were built by an Italian company that promptly went out of business. Replacements were impossible, spare parts scarce and technical help non-existent.

--The pedestrian route from the front of the bus to the sidewalk upstairs was, for a number of reasons, almost never a ‘straight shot’. The average time for an average person to get from upstairs down to the bus or vice versa was three minutes. In some stations, it took nearly five minutes to get from the station platform to the outside.

--Recently, while converting the downtown tunnel to light rail use, the buses were rerouted to the surface streets above. Third Avenue was effectively converted to a transit mall during rush hour. The buses using this transit mall were able to get through the one mile of downtown three minutes faster than the buses that had previously used the tunnel.

--And worst of all, the train tracks were installed incorrectly. They simply neglected to ‘ground’ the rails. Stray electrical current from the eventual operation of the trains would seek out the path of least resistance. This path of least resistance would be the rebar steel reinforcing in the surrounding concrete and would corrode that steel until it was weak and worthless.

This final mistake meant that when the light rail came, the track had to be ripped up and replaced (and since it was imbedded in concrete, that surrounding concrete had to be ripped up and replaced).

Now the track has been replaced and Sound Transit says that they will open the rail line including the bus/rail tunnel under the downtown in 2009.

--Sound Transit’s new light rail vehicles have a higher floor than the buses that had previously used the tunnel. They were faced with a choice of raising the platforms or lowering the floor. Though the cost for lowering the rail bed at the bottom of the tunnel was more, the board followed the recommendation of the staff and opted to dig out the bottom of the station floors.

Tunnels are not static structures. The circle of a tunnel is a series of stresses playing dynamically on one another. One cannot simply cut a substantial piece out of a circular tunnel and expect all to be well. Some on the staff warned of possible collapse if the bottom were cut out of the stations. They chose to ignore these warnings and proceed with lowering the floors of all of the stations.

But they have even worse problems in the tunnel and these are insurmountable.

Midway in the tunnel is a ninety degree turn under the intersection of Third Avenue and Pine Street. The tunnel route goes from east-to-west and turns to travel north-to-south. The buses, about sixty-five feet long, had a bit of trouble making this turn. The light rail vehicles are ninety feet long and each train consists of four of these cars. Sound Transit has a study from an engineer that says that they can make it. The engineer says that they have an inch or two clearance. They have pushed through one car, at three miles an hour. But four cars, going at normal speeds, will not consistently make this turn. They will scrape, hit and maybe jam in the tunnel. Maybe not the first day or even the first year, but someday, sometime, someone will go a little faster than posted, rock the train just an inch too far that way or this way and it won’t get through.

At best, they wreck a train. At worse, we don’t want to think about it.

As bad as that is, there is an even worse problem at the southern end of the bus/rail tunnel. Towards the southern portal, the tunnel goes under the freight tunnel built in 1905. The bus/rail tunnel heading south goes down a short but steep incline. The slope is 5.25 degrees, which means that it falls five-and-a-quarter feet for every hundred feet of horizontal distance. (The steepest non-cogged railroad in the world has a 5.5 degree slope.) Then it levels out for eighty-five feet and then continues up a short incline of 5.4 degrees. Sound Transit has also pushed a single car of ninety feet length through this stretch, again at three miles per hour. Again, it won’t be able to consistently get a four car train through at even reduced speed. Eventually, there will be a reckoning as Sound transit tries to force a square peg through the round hole of physics.

Neither of these problems is solvable. There is no way that the bus/rail downtown tunnel will be able to handle the new light rail train. At best, this will be obvious. At worst, it will be catastrophic. Knowing Sound Transit, I hope for the best and expect the worst. There will be a day when the train becomes a train wreck, wedged deep in the tunnel, underground and fatal.

When that happens, I would hope that the entire board of the Sound Transit resigns. They have been warned by better than me that this project was misdesigned, mismanaged and badly built. They should have done due diligence to avoid problems that will come their way. They should have taken care that things were done right. Instead, they ignored the warnings, turned a blind eye to the obvious and relied on staff recommendations—from a staff with a proven record of failure. They should resign as a matter of course. Then, they should resign from further public service, for they have shown that theirs’ is no service at all but a failure to meet even the minimum of standards. They should resign and hope to God that will be enough.

What we could do today, tomorrow and this year

I was speaking to a group of suburbanites against a ballot measure that would build more roads and transit (it was a proposition that costs 57 billion dollars, not finish a single road and the transit planned was worse than none at all). I gave my reasons why the proposal was so wrong and so on, and someone in the back yelled “Then what would you do instead?”

Good question. Here is my answer. You’ll notice that I will try to define the problem in a couple of sentences before I give the answers. Truly understanding the problem is the key to having the answers.

First, there is no way to build your self out of congestion. It won’t work. Cities like New York, Chicago and Las Angeles have tried it and failed. In fact, building more roads just creates more travel. Not only will there be more cars, trucks and other vehicles, but individual vehicles will travel more.

There are enough roads. It’s just that we all want to use a few of them at the same time. The average American city has paved at least 25% of its land surface for cars. Surely, if we spread out demand, there is enough concrete for all of us.

The second point is that 60% of congestion is caused by accidents, breakdowns and stalled vehicles. All of the roads in the world are useless if there are vehicles stalled, broken down or involved in accidents on those roads. You could do more to alleviate congestion by removing breakdowns and accidents than you can by building more capacity.

That being said, it is important to start working on solutions, both so that you can make some headway on the problem and to show people that you are at least trying. So the trick is to attack the problem from several fronts.

The first and easiest task is to spread out demand. The key is to flatten the demand curve as much as possible. This is not to say that you ignore the breakdowns and accidents. In fact, you need to do both remove the obstructions and modify demand. It just so happens that in the flip of the coin in my mind, demand modification won.

Change nearly everyone’s work schedule

Get rid of the 9-5, Monday through Friday work schedule.

Thirty miles north of Seattle are the huge Boeing 747, 777, and 787 plants. The assembly plants are about three miles east of the Interstate 5, the main north-to-south corridor. You leave I-5 and drive three miles on a spur freeway to the site. This roadway used to be jammed when the workers were released at the end of their shifts. Then Boeing got smart. Now, every worker has an individual, separate starting time—and ending time, of course. When the individual’s eight hour shift is up, they pack up their tools, clean up and head out. Their replacement (Boeing is working around the clock) comes in and takes their place. The result is that Boeing loses no production time as shifts change and the spur highway out to the Interstate 5 three miles away flows easily.

Obviously, there are people in the general workforce that must start at 9:00 and stop at 5:00 but they are a tiny minority. Most work is more flexible than that.

Newsweek is reporting that 27% of us will be telecommuting by 2010. Even if we aren’t all telecommuting, I would bet that many of us spend the first hour or two answering emails, taking care of the phone messages or just catching up on the news that affects our industry. This could be done from home.

Inform people of alternative routes

OK, so say that you are on the road and traffic begins to back up. Once again, I call on technology to the rescue.

We need to set up ‘traffic speed ahead’ indicators. These simple displays would consist of three numbers: the top number would be the average speed of the cars in the next mile, the second would be the average speed in the next five mile and the final number would give the speed of the next ten miles. These indicators would be placed at ‘decision points’. Let us say that you are approaching the on ramp for the southbound freeway into town. The indicator lists the speeds as “18”, “42” and “55”. This would tell you at a glance that there is a very bad problem within the first mile of the ramp, but that the rest of the freeway is moving well. You would skip this ramp, use surface streets to reach the next ramp south and use that access instead.

As Global Positioning Systems become more commonplace, they will improve. Truckers will not be afraid to try alternative routes even in unfamiliar areas. The GPS will guide them expertly around traffic. They will eventually be enabled to deliver exact alternatives with real-time savings displayed. By the end of the decade most new vehicles will come with GPS.

Clear the way during peak hours

There are some vehicles that simply should not be on the major highways during peak hours of traffic: mobile homes, huge ‘over-sized’ trucks, army convoys and a few others. There is no practical way of stopping them, but you could fine them if they get involved in an accident, breakdown or stoppage. Make the fine so high that no right-thinking company would risk it. Why in the world can’t they move the mobile homes between midnight and four AM; is someone waiting for them somewhere at 9:00 AM? (These vehicles are driven by pros and their record is probably better than most. That misses the whole point. It is not that these types of vehicles are involved in accidents and breakdowns disproportionaly, it is that when they are involved, it is a great big mess.)

If congestion persists, toll Single Occupant Vehicles off the road at peak hours

The political pressure to toll is becoming too great to resist. It can be turned to the advantage of those that want to reduce congestion. Nineteen out of twenty cars in the commute traffic carry only one passenger. The gains from carpooling are paltry. But if tolling was designed to ease congestion, we could simply not toll anyone with more than one person per vehicle. If the toll was set high enough to compel people to carpool with even just one additional person during the peak hours, we could see the total of cars cut nearly in half during peak hours. And that would be anything other than ‘paltry’.

Study where, when and who is causing accidents and breakdowns, and then go out there and stop those accidents and breakdowns

Transportation experts will tell you that this has been studied to death. They are wrong. The State of Washington does not keep a computerized list of all of the vehicles left at the side of the road. They do not have a law that if your vehicle is abandoned three times in five years, we keep the vehicle until you prove that it is road-worthy and won’t be left again (or you could sign an agreement not to drive on the freeways).

A few years ago, a new Chief of the Patrol in the state of Washington directed the troopers to concentrate on four types of violations and issue tickets. The Big Four were drunk driving, excessive speeding, aggressive driving and not wearing a seat belt. No more verbal warnings, no stopping to help someone change their tire, don’t sweat the tail light out, just going out and issuing tickets for these four offenses. Where they did this, the fatalities dropped by 25%. Some of the drivers got off the road. Accidents fell and traffic moved.

Tell your law enforcement people that you want the accidents to go down and you will back them to the fullest.

One way to do that is to enact laws calling for the confiscation of vehicles for some infractions. Drunk driving, no insurance and truly excessive speed should be punished not by fines or jail but simple confiscation of the offending vehicle. Nothing else will stop some people. Sixty days for the first offense, 180 days for the second and permanently—with crushing the vehicle the third and final time. This may seem excessive; we just had a driver with several DUI’s go the wrong way on the freeway kill someone in the car that they were unlucky enough to hit. Now, that seems excessive to me. Had we confiscated their vehicle a couple of DUI’s ago, this accident would have never happened. By the way, the accident closed the freeway for several hours. Given that someone died, this is not the biggest part of the problem. But if the vehicle had been seized, the closure of the freeway also would not have occurred.

Get law enforcement to video their arrests. Watch as the conviction rate nears 100%. Wisconsin did this and their conviction rate is 92%.

Generally, get serious about traffic law enforcement. Go after those things that cause accidents. Confiscate the vehicle of those people doing these things. Keep up the pressure and don’t accept accidents as a normal part of traffic. Yes, there will always be fender benders but the really big, traffic-clogging accidents are caused by really bad behavior and we can stop this. All we have to do is decide to act.

Have law enforcement get in front of vehicles and then, lead them off the road

Law enforcement pulls up behind the offending vehicle and them follows them off of the road where they write them up a ticket. There are three things wrong with this out-dated approach.

First, the presence of the cop car does slow down the other cars on the road, but to some degree, they slow down too much. The cars behind them, seeing tail lights, slow and so on and pretty soon, you have traffic jams.

Second, too many cop cars are hit while sitting at the side of the road. Now, it is mostly drunks slamming into the back of the cops but this is a lousy way to catch drunk drivers. In the past year, seven state troopers have been hit at the side of the road.

Third, if the police pulled in front of the offenders, activated a lighted sign on their trunk that said something like “Slow down, follow this vehicle off of the road” with blue lights and sirens for added emphasis, this would virtually eliminate high speed chases. (The drunken driver that I cited above driving the wrong way on the freeway was being pursued by a police officer. Had the law enforcement vehicle been in front, there would be no chance that the drunk driver could have entered the freeway on-ramp the wrong way.)

Clearing accidents and breakdowns should be, after taking care of the injured, the highest priority

Imagine if I dropped a load of wood onto the roadway. Or someone had a mattress fall off their truck. Would we allow the obstruction to stay there? Of course not. When someone is involved in an accident without significant injuries, we should simply push the vehicles off of the road. Any additional damage that the vehicles suffer would simply be added to the damage caused by the accident. It is maddening to see literally tens of thousands of people stopped simply because we do not feel that we can order the state troopers or city police to push the cars off of the roadway.

The shoulder lane of the highway is the most important lane of any highway

In Seattle, in our great rush to build High Occupancy Lanes, we decided to use the shoulder lanes as HOV lanes. This was a giant mistake. The shoulder lane is where you take your accidents, where police write tickets and where we can safely put breakdowns. It is the lane that keeps the others—all of the others—flowing freely.

The shoulder lane should be considered sacred and treated as such.

Bring back the shoulder lane. Even though it would technically reduce capacity, it would substantially increase flow on all lanes. It would reduce congestion. Even if we had to forego HOV lanes, flow would be improved.

Design for flow of traffic

The trick is not to build more lanes of traffic but to use the lanes that you have to better effect. Cars stopped are lanes of traffic that do not work—in fact, the lane might as well not exist at all. Keep things moving on the lanes that exist and you won’t have to build more lanes.

There are the accidents and breakdowns, but there are also the single car parked in the curb lane after three o’clock in the afternoon that effectively reduces the lanes available by half. There are the traffic signals that are holding up thirty-five cars and two buses for no good reason because there are no cars coming in the other direction. There are too many entrances and exits on the old Highway 99 that make that six-lane road both slow and dangerous.

This is not to argue for all-flow, everywhere. The neighborhood streets should be flow-restricted. They should not be used as shortcuts for people and things going from Rainier Valley to Beacon Hill, from Greenwood to Wallingford and downtown to Capital Hill. But the major highways, the arterials and the freeways ought to be made free flowing.

Modernize your traffic control lights to move traffic—and for crying out loud, add ‘left turn only’ arrows. Restrict parking and treat blocking parked cars like the obstructions that they are (you don’t need to tow them across town; just take them around the corner or nearby and put a lock on the wheel). Go to where the chokepoints are on major throughways and eliminate those problems.

Change the laws—especially tax laws—to favor driving less, or not driving at all

Every law, every rule ought to be changed to always make it better to drive less or not drive at all. Right now, there is no real change in the cost of insurance (the largest single check that you write for the privilege of driving) if you drive fewer miles or don’t drive one day. Make every insurance company offer a policy that notes how many miles you drive and lets you pay by the mile. Better, consider making minimum insurance part of the cost of a gallon of gas. You buy a gallon of gas and twenty-five cents of the price buys you collision insurance (covers the other person in case of accident). Don’t worry insurance companies, there is still plenty for you to sell in the form of extra added coverage that most car owners will want. But if you don’t drive, you don’t pay for insurance.

We should help those who choose not to drive or drive less. The state could try tying your cartabs to the amount that you drive instead of the cost of the vehicle. The local cities could pay the sales tax on bikes. State and local governments could simply give people money—or more likely bus passes, rebates on bikes and the like—for not owning a car.

I know that some of us need to have and drive a car (I consider myself one of them). But every rule, law and tax ought to reward you for not driving or at least driving less.

The reason for this is simple: everything that you do to releave congestion opens a place and someone on a bus, in a carpool or who has switched to off-peak hours, says “Hey, I can drive alone at 8:30” --and does. Unless and until you get the total number of vehicles off the road, you are really just spinning your wheels.

Sisyphus rolled up the rock to the top of the hill every day just to see the rock rolling back down the hill, to begin again rolling the rock up to the top of the hill again. Then one day, Sisyphus said to himself “Hey, this is stupid” and stopped rolling the rock up the hill and then the rock didn’t roll back down and he went and started a dating service for the Gods on the Internet and made a million bucks and retired to Montana or something very like that. The point is, if you keep rolling a rock up the hill and the same thing happens every day, you really should try something else. I don’t have the answer, but I think that we ought to try lots of things, see what works, discard what doesn’t and stop doing what is obviously not working.