Monday, September 18, 2017

Is It a Curve or Is It a Turn?

Making the turn, child in tow.
I've been looking at the intersection of 27th Street and Lombard Street in Philadelphia. It's a key part of the westbound access to the South Street bridge. The one thing I know for sure is that it is not an intersection. There is no westward extension of Lombard, past the intersection; and there is no 27th Street north of the intersection. 27th here is a one-block stub leading south to the bridge. There's another one-block stub of 27th just the other side of South, which feeds cars onto the bridge from Schuylkill Avenue.

Lombard and 27th are effectively one street that bends awkwardly at approximately 90 degrees, at the point where the nomenclature changes.

So this 90 degree thingy, is it a curve, or is it a turn? The people who designed and built this little stretch of road seem to have been genuinely conflicted by this question.

Lombard and 27th Street come together.
Here's another way of expressing the dilemma: Is it a city street, or is it an access ramp to the bridge and the Schuylkill Expressway, aka I-76?

Points in favor of ramp. There's no traffic light at Taney, the cross street just before 27th. (I think the ramp vibe starts at Taney.) There is also no traffic light at 27th, where there is, admittedly, no cross traffic. But the lack of a signal at these locations is a signal.

As you come to the turn, you'll notice the designers have gone to considerable effort to at least make it look like it's not a full 90 degree turn. The road widens substantially at the corner and the curbs don't form right angles, but instead present gentle, wide curves. On the inner side of the turn, this effect is enhanced by the judicious use of paint. All this encourages people to act as if they're swinging along on an interstate access ramp.

Finally, there are no crosswalks at Lombard and 27th. I personally think you'd have to be insane to try to walk across the street here, crosswalks or no crosswalks. But it's another little clue that this is not a city street.

(There are no crosswalks across Lombard at Taney either. People do walk across the street in this area with some regularity. Remember, there are lots and lots of people walking on the bridge, and they have to get there somehow.)

Points against ramp. If the turn at Lombard and 27th actually was an interstate access ramp, the curve would be banked.  A banked roadway makes it easier for cars to turn; it also means that all drainage goes to the low side of the bank, which in this case would be the left side of the road.

Instead, the street at this point has a crown in the middle and drains to both sides. This means that people on the right side of the road are turning on a surface that has reverse camber.  The problem with reverse camber in a turn is that it tends to throw you off the road. Which is why curves are generally banked.

Not surprisingly motorists tend to steer through this area slightly to the left of the crown, where the camber helps them turn. When they do this they need to avoid a large storm drain located in a depression in the pavement. They can do this by going to the right of the storm drain, or by straddling it. You don't want to put a wheel in that depression.

A storm drain for the motorists.
My friend Bill Marston thinks the drain probably started life next to the curb, but then the curb moved several feet closer to the building. If we accept this line of thinking, the streets engineers wasted their time moving the curb, because the bulk of the traffic is going to the right of the drain, and the motorists who are straddling could easily move to the right. So you have the appearance of a wider street, but not the reality. (The gap between the curb and the grate is 6' 6". I measured it.)

What's going on? I think the cartway's profile here is driven less by the needs of drivers and more by some thorny issues of drainage. The intersection of 27th and Lombard is at the bottom of two hills, one running down Lombard and the other coming down 27th from the bridge. When it rains, this intersection is definitely collecting storm water from a pretty wide area.

In addition to watching motorists, I've been watching bicyclists navigate through this area. They're hardly ever in the bike lane at the corner. They're to the left of it, I think for two main reasons: First, the higher route allows for a gentler curve. Second, there is a fearsome storm drain designed to catch the wheels of bicycles and eat them, and it's located at the curb in a particularly infelicitous spot. (I've also heard comments about gravel gathering in this spot. I wouldn't be surprised, since it's at the bottom of two hills. I just didn't see it.)

Storm drain in the bike lane.
I don't have solutions for any of these issues, but as we redo the bike lane in this section, I just wanted people to be aware of some of the design challenges.

Riding the curve.
See also Morning on Lombard Street, No Turn On Red, Put Traffic Lights on the Schuylkill Expressway.

Monday, September 11, 2017

My Life in Fairmount Park

Stas in the Vltava River. Prague, 2013.
I've run ten marathons. Paris twice, New York three times. I trained by running in Fairmount Park. I would run out Kelly Drive to the angels or the Kelly statue and back to my house on Lombard Street. Or I would run the loop out Kelly, across the Falls bridge and back on MLK Drive. That was the counter-clockwise route. Sometimes I would run clockwise. Sometimes I would do two loops. And sometimes I'd run out MLK Drive to the water pumping station on Montgomery Drive, and then over to the Belmont Plateau for some hill work. Nice view from the top, if you still had binocular vision.

A significant component of my training was on a bicycle. When I was too tired to run, I would bike. I have happy memories of MLK Drive, and the lights under the Strawberry Mansion bridge, before dawn on a February morning. There was nobody else there. I felt safe, and I was happy, even though parts of me were quite cold.

The only part I didn't like was the crossing of the bridge at the beginning of MLK Drive. It was acceptable when I was running, but when I was biking I always knew fear.

I started writing about biking in Philadelphia in 2012. I wrote about the MLK bridge. I approached various people. I spoke with my City Councilman's chief of staff. I have subsequently spoken with many powerful people.

The MLK bridge looks just as it did in 2012. Five years have passed, and many wonderful things have happened in my life, including the birth of my first grandchild. But nothing has happened on the MLK bridge.

That's not my fault. I've done my bit. The failure lies with our city fathers and mothers.

Vaclav Havel. Prague, 2013.
"Hope is not the conviction that something will turn out well but the certainty that something makes sense, regardless of how it turns out." - Vaclav Havel

See also Uncorking the Bottleneck and The Bottleneck on MLK Is Still There.

Tuesday, September 5, 2017

Furness Buildings in Rittenhouse

Well Known and Lesser Known

Bates house, 1801 Delancey, 1867-1868.
I'd been walking by the building pictured above for decades, but it was only recently that I discovered Frank Furness had had a hand in its design. Which explains a lot about the Bates house at 1801 Delancey Place, on the northwest corner of 18th and Delancey.

As Michael Lewis puts it, "The house was an unconventional townhouse design for Philadelphia, where custom placed the rowhouse to the front of the lot, leaving a space for a private yard to the rear. The Bates design reversed this: a garden was created in the front, screened by a brick wall, a very eccentric feature." The result was "an unexpected suburban enclave on a street of stiff late Georgian townhouses." (Michael J. Lewis, Frank Furness: Architecture and the Violent Mind, 2001, p. 63.)

Not that Furness and his firm were averse to working on the gorgeous Georgians of Delancey Place. Records indicate the firm did alterations in 1820, 1821, 1823, and 1830 Delancey.

At the other end of the 1800 block of Delancey, just across 19th Street, is the famous Horace Jayne house, from much later in Furness's career.

Horace Jayne house, 1900 Delancey, 1895.
I'd been aware of the Jayne house and the Thomas A. Reilly house up on Rittenhouse Square. (The neighbors are having some work done.)

Thomas A. Reilly house, 1804 Rittenhouse Square, 1891-1892.
But mostly I'd been aware of Furness as a posthumously tragic figure whose unique style went out of style and whose buildings had an uncanny affinity for the wrecking ball, particularly in the years just before the rise of the historic preservation movement in the 1960's.

I'd never really looked into Furness. However, when I was working on my story about creating a large piazza to the west of Philadelphia's City Hall, I found myself checking books on Furness out of the library. I got what I needed out of them, finished the story, and then I kept reading.

A lot of Furness's best stuff did get hammered into dust and chunks, particularly the banks down on Chestnut near Independence Hall. It's a shame. We could be reusing those banks now for restaurants and museums. (Museum of the American Revolution, anyone? I guess we missed our Musee d'Orsay moment.)

I think there's a subtext, even today, to the conventional wisdom about Furness - that his buildings are so quirky they can't possibly work well. And that take is a mistake. Furness was, among other things, a master at moving people through space; his Broad Street Station, which used to stand just west of City Hall, was a prime example until it was knocked down.

A number of masterpieces remain - particularly the Pennsylvania Academy of Fine Arts on Broad Street, north of City Hall, and the Furness Library out at Penn. And, in the area around Rittenhouse Square, where I live, it turns out that we have a substantial number of surviving Furness buildings.

Some of these are institutional: His dad's old congregation, the First Unitarian Church, and his addition to the building that housed the Pennsylvania Institution for the Deaf and Dumb, now Dorrance Hamilton Hall of the University of the Arts. Hamilton Hall, by the way, is across the street from Furness's childhood home at 1426 Pine Street.

But most of what we still have in the Rittenhouse area is residential (we lost two major train stations, the Pennsylvania Railroad's Broad Street Station near City Hall and the Philadelphia Depot of the Baltimore and Ohio Railroad, 24th and Chestnut).

I'd heard the Thomas Hockley house was by Furness, but I didn't really know. It is. When you think about it, the corner entrance, the pigmy columns, and the chimney that leaps out of the wall are certainly hints.

Thomas Hockley house, 235 South 17th Street, 1875.
A bunch more, almost all of them buildings I have walked by many times, I simply had no idea were by Furness.

I'm not even going to try to list them all - it would be too long a list, and I know I'd miss something - but here are a few Furness houses that can round out a walking tour of Furness in Rittenhouse.

Let's start up on Walnut Street, with the John Rice rowhouses. 2106 on the left has fared better than 2108 over the years. 2108 is currently undergoing a vigorous gut rehab, but it appears that what is left of the historical facade will be preserved. And then there's that new construction to the right. We can hope for the best.

John Rice rowhouses, 2106-2108 Walnut Street, c. 1870.
Next we can hop down to the E.B. Warren rowhouses at 2102-2106 Spruce Street.

E.B. Warren rowhouses, 2102-2106 Spruce Street, c. 1871.
A little while later the firm supplied a similar facade for the Rudolf Ellis house at 2113 Spruce. Having some work done here as well.

Rudolf Ellis house, 2113 Spruce Street, 1873.
On 17th Street north of Walnut, we have a more commercial neighborhood and a nice row of buildings known as the Caroline Rogers houses (124-132 South 17th Street). Originally there were five, but the northern two were lopped off, and now there are three; 128 is no longer recognizable as a Furness building, so that leaves us with 130 and 132, and they could use some work. The iron beam over the doorway is echt Furness.

Caroline Rogers houses, 124-132 South 17th Street, 1887.
If you're interested in exploring further, allow me to send you to George E. Thomas et al., Frank Furness: The Complete Works, 1996. The catalog there supersedes the checklist in James F. O'Gorman, The Architecture of Frank Furness, 1973, which nonetheless remains a valuable resource.

I need to warn you that going through the catalog is a bit like reading the casualty lists from the Battle of the Somme in World War I. So much has been lost.

Let's go back for a minute to the little Thomas A. Reilly house at 1804 Rittenhouse Square. It had a big brother next door, the William West Frazier house at 250 South 18th Street, built in 1881-1882. It occupied the area from 1804 Rittenhouse down to the corner of 18th Street, and the front door was actually on 18th.

This site is now occupied by a large apartment building of which I am rather fond. But along with my fondness for the present, I must recognize that we lost one of Furness's signature works here. As Thomas et al. note (p. 230), "This was one of Furness's most important commissions located at the corner of Rittenhouse Square - one of the most visible sites in the city and Philadelphia's premier square."

I spend a fair amount of time in Asbury Park, a city which has also seen its share of devastation. A few years ago some intrepid techies put together an app that provided 3D images of structures along the boardwalk that aren't there any more, or that have changed greatly. As you walked along the boardwalk with a tour guide, you pointed your cell phone or tablet at a site, and up popped a ghost building. For good measure you could see the SS Morro Castle where it ran aground next to Convention Hall in 1934.

The app is called Augmented Asbury Park, and although there are no more walking tours it appears there is an online version.

My thought is that some group of intrepid techies in Philadelphia might like to do a similar app for missing Furness buildings. The south side of Rittenhouse Square would be a good place to start. Call it The Furness Ghosts.

This chimney at 132 South 17th has seen better days.

Monday, August 28, 2017

No Turn On Red

Vision Zero Meets the South Street Bridge


Sunset Avenue Pavilion, Asbury Park.

Worst things first. The intersection at 27th and South, at the foot of the South Street bridge, presents ... some issues. The worst of these is the traffic that comes off the bridge and turns right to go on to Schuylkill Avenue. I understand that the City's traffic engineers want to get these darling motorists home and in the arms of their loved ones as soon as possible. But perhaps it would be wiser - shall we say more balanced - if they tarry at the light for a few seconds. There's a lot of pedestrian traffic on this bridge - much of it crossing with its back to the turning traffic. And there are various vehicles - both motorized and unmotorized - that come down 27th, cross South, and seek to go down Schuylkill Avenue.

I was there Thursday afternoon, August 24, and I saw three near misses in one hour.

If Vision Zero means anything in this city, it means a No Turn on Red sign at this intersection. Not next year. Not sometime before the next mayoral election. Right now. Make the call, Jim Kenney.

Okay, so let's back up. Why was I out there? Well, a little while ago I posted an article about the westbound traffic on the bridge, suggesting that the traffic that backs up Lombard Street in the morning rush might be alleviated by putting traffic lights on the Schuylkill Expressway. And it got me thinking again about the eastbound traffic. My main concern for the last few years has been the garage entrance for the new CHOP building.

No longer. The worst spot eastbound is the intersection at the eastern foot of the bridge.

Early Friday morning, the South Street Bridge at 27th.

There are other issues, but again, let's back up a bit. I went to the bridge three times in two days and did traffic counts.

A Few Surprises
On Thursday morning, August 24, between 10:45 and 11:45, there were 439 motor vehicles in the left lane crossing the intersection and proceeding east on South Street. In the right-turn lane, headed for Schuylkill Avenue and points beyond, there were 178 vehicles. Of these 178, ten changed their minds at the last minute, swerved across the bike lane, which at this point lies between the two car lanes, and proceeded eastward down South. (Total cars = 617.)

On Thursday afternoon, between 4:50 and 5:50 p.m., there were 370 vehicles in the left lane, heading east on South. There were 301 vehicles in the right-turn lane, heading to Schuylkill Avenue. Of these 301, 13 changed their minds and jumped over to South Street. (Total cars = 671.)

On Friday morning, August 25, between 7:50 and 8:50, 410 vehicles used the left lane to get to South Street. In the right lane there were an additional 201, with ten of those bolting to the left and proceeding down South. (Total cars = 611.)

The totals for each hour are similar, but the composition varies. The highest number of vehicles proceeding down South was on Thursday morning; the highest number of cars using the right lane was on Thursday afternoon.

My main learning here is the number of cars switching from the right lane to the left, and crossing the bike lane to do so. This follows the crossover, where the bike lane moves to the left and the right-hand motor-vehicle lane moves to the curb. This crossover is challenging in itself, but at least people are aware that it is going to happen. What they are unlikely to anticipate is that, approximately every five minutes, a car will cross the bike lane to get from the right lane to the left.

I think the current lane configuration at the east end of the bridge is fundamentally flawed, and not fixable by palliative measures. We need to see people as they are, not as we would have them be, and then we need to design accordingly. A human factors engineer in a good mood could write a very amusing report about this intersection in its current state.

Also observed but not recorded were several u-turns, a number of drivers violating the bike lane near 27th by moving from the left lane to the right and then proceeding to Schuylkill Avenue, and a number of cars stopped in the bike lane. It might be helpful if CHOP management urged its employees to pick up their Uber rides on Schuylkill Avenue, and not on the bridge. Finally, traffic in and out of the CHOP garage was light at all times.

Kill the Turn Lane; Add a Lane Westbound
My initial thought in doing these traffic counts was a desire to free up space for an additional westbound lane on the bridge. As it now stands, the westbound traffic on the bridge starts in one lane, which eventually blossoms into three. If westbound traffic backs up into the single-lane area of the bridge, which it frequently does in the morning, then traffic can quickly back up Lombard as far as 22nd Street. Adding a second lane here would allow westbound and southbound traffic to flow through, avoiding the queue for the northbound Schuylkill ramp.

I don't think adding the second westbound lane would solve all the problems here - the basic problem is the Schuylkill Expressway - but I think it would be a substantial help.

So should we kill the eastbound turn lane and add a westbound lane? Well, I'm for it. It would allow the eastbound bike lane to stay at the curb, avoiding both the crossover area and the bandits who violate the bike lane near the intersection. A No Turn on Red sign would make life much easier for the many pedestrians on the bridge, and overall the intersection should become much calmer. And the snake of traffic that we see so often on Lombard Street should become much shorter and appear less frequently.

The downside is that eastbound traffic will back up much further than it currently does. My observations at the end of August revealed plenty of back-up room on the bridge. It's true that traffic will be heavier after Labor Day. Perhaps someone who gets paid to do this stuff would like to go out and see if numbers from the fall invalidate my basic thesis, which is that there is room on the bridge to queue more cars, and that the bridge is a more appropriate location than Lombard Street.

Beyond that, however, is the question of safety. The current mayor has announced his commitment to Vision Zero. I believe my proposed configuration on the bridge would substantially increase safety for all bridge users. So, tell me how keeping things the way they are fits in with Mayor Kenney's commitment to Vision Zero.

See also Put Traffic Lights on the Schuylkill Expressway and Morning on Lombard Street.

Tuesday, August 15, 2017

Broad Street Cycle Track

This was originally posted on the Facebook page of the Bicycle Coalition of Greater Philadelphia on June 4, 2014:

Just to be truly radical for a moment, what about a cycle track on Broad Street in the median? We could also add parking lanes next to the cycle track, so that vehicular traffic would be reduced to one lane north, one lane south. These parking lanes would become turn lanes at intersections. The parking lanes lanes next to the sidewalk could then be more heavily dedicated to delivery and drop-off.

Sunday, August 6, 2017

Put Traffic Lights on the Schuylkill Expressway



Waiting patiently for a green light on the South Street bridge. 

This idea came to me on July 6 of this year. It had taken me a while to get there. Last year I wrote a story called Morning on Lombard Street, which described the hellacious mess that Lombard Street can become during the morning commute.

On a bad morning, traffic can back up Lombard to 22nd. It turns out that this congestion has essentially nothing to do with the bike lane, although people continue to mention that possibility to me. The congestion is caused by the Schuylkill Expressway. When it backs up heading north, the northbound access ramp from the South Street bridge backs up and then the bridge backs up. If it's bad enough the backup goes across the bridge and then snakes back up Lombard.

If you'd like more of the nitty-gritty on this, read Morning on Lombard Street.

I'd never actually taken pictures of these backups, so last Friday, August 4, I went out in the morning to snap a few shots. It being a Friday morning in August, traffic was very light, and I found myself more interested in another phenomenon that many people seem blissfully unaware of. Cars do not own the South Street bridge. There are scads of bikes and pedestrians and dogs. I'm actually thinking of setting up a coffee stand on the bridge, just next to the northbound ramp to the Schuylkill. And I'll definitely include a free doggie water bowl.

On July 17 there was a community meeting about upgrades to the Lombard and South Street bike lanes. I'd been aware that there was a lot of bike traffic on the South Street bridge - Lombard feeds the westbound bikes onto the bridge, and South takes the eastbound bikes off the bridge - but I was actually surprised to learn that 15 percent of the vehicular traffic on Lombard is bicycles. Then you need to add in the pedestrians, and of course the dogs.

The South Street bridge is not an urban wasteland of concrete and cars.

Six people and a dog in search of a green light.

Even though the car volumes weren't there on Friday, it's interesting that the northbound access ramp still accounted for the lion's share of the cars, as it does on heavier days. The picture below gives you the idea. Westbound and southbound are running clear. The northbound ramp traffic is backed up almost to the CHOP building.

Do not take pictures while walking in a crosswalk.

Here's another shot, showing the Achilles heel of the whole thing. If the cars back up any further than this, you're into the area where there's only one lane. And that means that all the westbound and southbound traffic gets snarled up with the northbound traffic. And some days the snarl goes back up Lombard to 22nd.

The Achilles heel.

Last year, in Morning on Lombard Street, I recommended closing the northbound ramp. Among other things, it's a very dangerous place. You need to enter the Expressway in the fast lane, and the sightlines are not good. People with long experience in this matter told me that my proposal was not new and would never happen.

So here I am with another idea. I recently read that Market Street and JFK Boulevard were designed as "urban highways." They have traffic lights. Instead of treating the stretch of the Schuylkill around the South Street bridge as an interstate, let's call it an urban highway. It doesn't have cross streets, but it has scads of exit and entrance ramps, and the access ramps are pretty much all problematic. Let's put lights across the main roadway and also at the front of each access ramp. Then let people take turns. I'm thinking things will go  better.

When traffic is heavy, something close to this already happens on an informal basis. The cars in the main roadway slow down, and every once in a while they actually let someone in from one of the access ramps. I'm just suggesting that we use familiar, approved traffic control devices to formalize this dance, reducing the danger and frustration that are endemic to the current arrangement.

Waiting at the light by the Schuylkill access ramps.

Let me close with one last picture - I am a bicycle guy after all, and I believe there are more bicycles than cars in this picture. I've been waiting three years to see that. The future is here. It's on the South Street bridge. We just need our village elders to snap out of their Rip Van Winkle act and help us build upon what we can already see with our eyes.

See also Intermittently Terrifying.

Wednesday, August 2, 2017

Balkanized Politics

The Dependent Probability Trap

Esplanade by Fairmount Water Works
I suspect that the balkanized politics of Philadelphia has its basis in neighborhood resistance to urban renewal. What better way to stop unwanted change than to devolve power from City Hall to a series of small community organizations, any one of which can stop any change.

I'm afraid I don't know enough Philadelphia history to be able to prove this case; perhaps someone else can do that. What I do know is that Philadelphia politics is currently well designed to stop just about anything from happening.

And I'm going to show you how it works. Say you have a bike lane that is essentially the backbone of a citywide network of bike lanes. Say it runs through the bailiwicks of three registered community organizations. Say you want to make some improvements to the bike lane.

And here's what happens. The local city council members put their heads together and decide that they're okay with the plans, but that all three RCOs need to hold community meetings and get neighborhood approval. Everybody has to agree.

Let's say that your chances of RCO approval are actually pretty good - say 0.8 on a scale of zero to one.  But your job is to get all three RCOs to agree. And here we run into what I call the dependent probability trap. The probability of getting all three RCOs to agree is not 0.8. It is 0.8 x 0.8 x 0.8. Which is 0.51, or fifty-fifty. Add one more RCO and it is 0.40.

The only way out of this trap is for our elected officials to engage in something called leadership. They need to stop giving their constituents a veto. They need to listen carefully, take counsel, and then make decisions.

It's a much harder job than just ducking for cover anytime a decision looms on the horizon. But it is the job we pay them to do.