Bike commuting in DC highest in US Northeast - among highest in nation

According to a recent article in the The Atlantic Cities, bike commuting has seen significant increases across the US between 2000 and 2009.  In the Northeast region, DC led the pack, tied with Philadelphia, at 2.2%.  That is up from 1.2% in 2000.
Although DC's increase in cycling was not as remarkable as some of the other Northeastern cities, DC started the decade with more bike commuters.

My Commuter Mug Reprise

photo by Scott Feldstien on flickr

Long ago I posted that by carrying my commuter mug around with me I would save money on my coffee purchases.  Starbucks always offers a dime and sometimes they just charge me $0.50 for a refill even though it isn't a "refill."

This morning I was in Shirlington and dropped into the Caribou Coffee shop.  They offer a $0.50 discount for bringing your own mug.  Wow!  And, it was prominently displayed on their big chalkboard.

I recently blogged on how Starbucks has failed in their efforts to improve use of reusable mugs.  Well, Caribou is kicking their butt!!

I-66 "Spot Improvement" Underway

The "spot improvement" on westbound I-66 between the George Mason Drive and Sycamore Street is underway.

These spot improvements, which are essentially short-distance widenings have been the subject of great debate for more than a decade.  VDOT has long wanted to widen I-66 to three lanes in each direction; Arlington, smart growth and environmental groups have been strongly opposed.  The original agreement to build I-66 included an agreement to not widen the highway beyond two lanes in each direction.  VDOT has instead proposed a series of "spot improvements," which widen sections of I-66 to three lanes.  This section is called "Spot 1," and it has a page on the VDOT web site.

In February of 2009, the National Capital Region Transportation Planning Board voted to put these projects on hold,only to reverse itself the very next month.  Since then, there has been little to stop this project from moving forward, particularly because it has received earmarked funding from Congress that can be used for nothing else.

The contract was awarded in May, and construction has begun.  Here are photos:

I-66 looking westbound from Ohio St.

This one shows I-66 looking westbound from Ohio Street.  Note the (semi) vegetated median between the left shoulder and the Metrorail tracks.  That is being partially removed to make room for the additional roadway width.

The photo at the top of the post shows the demolition of the median next to the Metrorail tracks.

The Custis Trail passes under I-66 along this stretch of highway.  At the point were it crosses underneath there are three bridges: one for each direction of traffic and one for the Metrorail tracks.  The westbound bridge is being widened.  To accommodate pedestrian and bicycle traffic during construction, VDOT has constructed this structure for protection.  There have been occasional detours while work is being done here.  The detour is well marked and only slightly longer.

Once the bridge is widened, there will be less daylight at this point.  Although there is some relatively inadequate lighting now, it may need to be upgraded as part of the improvements.

The following information was mined from the project documents of the Constrained Long-Range Transportation Plan for 2030:

Do traffic congestion conditions necessitate the proposed project?  Yes
If so, describe those conditions: Recurring congestion
Is this a capacity-increasing project on a limited access highway or other arterial highway of a functional class higher than minor arterial? Yes

Please identify any and all planning factors that are addressed by this project:
  - Support the economic vitality of the metropolitan area, especially by enabling global competitiveness, productivity, and efficiency.
  - Increase the safety of the transportation system for all motorized and non-motorized users.
  - Increase the ability of the transportation system to support homeland security and to safeguard the personal security of all motorized and non-motorized users.
  - Increase accessibility and mobility of people and freight.
  - Promote efficient system management and operation.
a. Is this project being proposed specifically to address a safety issue? Yes
Briefly describe the nature of the safety problem:
Existing levels of congestion is exacerbated by the intense weaving and merging movements happening over a short distance along with inadequate sight distance. The recurring congestion and associated operational/safety effects poses concerns on the corridor’s ability to serve as an efficient emergency evacuation route.

The suggestion that this will improve the road as an emergency evacuation route is just silly.  The capacity of I-66 for evacuation will not be increased by adding a lane for a mile and a half.  This language is likely a holdover from the arguments used for widening the entire length of I-66 inside the Beltway.

Many people believe that once the "spot improvements" are all in place, the argument to widen the relatively short remaining sections will grow very strong, and VDOT will eventually prevail.  It will be interesting to see how long that takes.

(Cross-posted on Greater Greater Washington)

Congestion Charge, Commuter Tax and Metro

(Rerun - Originally posted Nov. 17, 2007) Last week a Washington Post reader named Bill Suffa wrote a letter to Dr. Gridlock in which he suggested that:
1) a congestion charge is essentially nothing more than a commuter tax and,
2) that Metro would be incapable of handling the extra ridership that would be created if a congestion charge for the district (or a portion of it) were enacted.

On point 1, I suppose he's essentially right, although "commuter" would include anyone who enters the congestion zone regardless if they are a resident of DC or not. My personal opinion is I don't see a problem with that. If the District wants to charge people for using its infrastructure (roads, bridges, etc.) then why not? Would it be good policy? Personally I think it would be, but others will disagree. There would be trade offs between raising money and driving away business, maintaining and improving infrastructure and irking suburban commuters, environmental advantages to reducing car traffic and other benefits and costs. I don't pretend to know what the right balance for the District would be, but I do maintain that they should have the right to decide for themselves, and I would suggest that if Fairfax County or any other jurisdiction wanted to charge for the use of its infrastructure, too, it should have that right.

On point 2, the challenges facing Metro have been written about a lot. Just recently it was reported that ridership has grown by 70,000 riders per day over the last 5 years (WMATA press release). Crowding on the system is getting worse despite new rail cars and eight-car trains being introduced. Mr. Suffa has experienced this himself, as have thousands of other riders. I doubt anyone would disagree with the statement that Metro service can be improved and capacity needs to be increased.

London put #'s 1 and 2 together. A significant part of the reason London was able to enact a congestion charge was that they simultaneously invested more money into buses and trains to make it easier to travel in London without having to drive in. It was part of the plan.

We can take a lesson, too, from the new Nationals Stadium. DC was able to float more than $600 million in bonds to finance the stadium from taxes that haven't yet begun to be collected and will continue for many years. London collects an estimated $244 million per year from its congestion charge and estimates for DC start at $60 million per year. Using similar financing rates to the stadium deal (and assuming between $60 million and $244 million per year were collected), if the congestion tax were dedicated to WMATA, bonds could be floated for between $900 million and $4 billion--money that could be put to work immediately to make significant improvements. Using the stadium strategy would allow for starting the improvements in advance of instituting the congestion charge, increasing the capacity needed before it is required.

New Model for Transportation - Take 2

My post from last week: We Need a New Model for Expanding Transit received a long, thoughtful response from Paul Minett: Co-founder, CEO and President at Trip Convergence Ltd. He was responding not to my post here but to my sending similar thoughts to an industry list serve called transp-tdm (TDM stands for Transportation Demand Management, and there is a whole industry of professionals who work in this field). I have edited Paul's response slightly, but have hopefully captured his thoughts and ideas:

Well said, Steve. And since you asked, here is what I think. I think the score card you propose is all wrong. This is not (or shouldn't be) a 'cars vs transit' issue, nor should it be a 'vehicle miles per gallon' issue. Eventually we will be forced to restructure how we achieve 'accessibility', and how much 'mobility' we actually need. In the meantime switching cars for buses or trains is like the proverbial 're-arranging the deck chairs on the Titanic': eventually they all become obsolete.

It has been suggested (by Jonathan Levine, for example) that 'accessibility' is made up of 'proximity', 'connectivity', and 'mobility'. We get what we focus on, and while the focus is on 'mobility' that is what results. A concerted effort to improve 'connectivity' could reduce the need for so much 'mobility'. On the other hand, within 'mobility' sits vehicle occupancy.

It amazes me that when some people look at a jammed freeway they see the solution being to add more vehicles (ie buses). There is enough rolling stock already out there. And it could all be getting 100 passenger miles per gallon, just by filling up the empty seats. We recently got almost 200 passenger miles per gallon on a trip from Portland to Salem: four people in a Prius. Eventually we must reduce the amount of 'mobility' we consume.

In the meantime we need to put lots of thought (and action) into strategies that increase average vehicle occupancy, and reduce total vehicle counts, and reduce the amount of 'solitary mobility' we consume. Possibly the most energy efficient transport available is the van owned by a worker who carries eight or ten co-commuters for most of its journey, and then parks up while the worker does his job, until doing the return journey.

Highly efficient buses that deadhead empty to collect the next load are only half as energy efficient on a per passenger mile basis as the same buses operating full in each direction. During peak the loadings might justify the deadhead, but outside of peak it becomes questionable.

As a starting point we could all focus on mechanisms to increase the rewards for sharing rides, increase the cost of driving alone, and simplifying the process of sharing rides. We are one of many businesses focused on finding solutions to make progress in this area, and it is not plain sailing. Many seem to be offended that private companies might make a profit in this space. But perhaps more to the point, we find a lack of discretionary time and budget in the offices of the organisations that should be interested in working with us. And when there is funding for innovation the processes to bring it about are complex and very slow. Only the most dedicated stay the course.

I think it is great that you have opened up this discussion. I hope others will find time to comment because this is a fundamentally important issue.

Thanks, Paul, for your ideas and helping to push the transformation to sustainability this blog and millions of others seek.

16? 2? 53A? &%@**#!!

IMPORTANT NOTE: In this post I may use ART and Metro as examples. That's because--being in the DC area--I know more about them. The points I am making are intended to apply to mass transit in general and are not intended as criticisms of these particular systems any more than transit systems anywhere else.

In Arlington (and many places) streets are numbered, which can be pretty useful: 10th St. is between 9th and 11th, which can be a huge help to those trying to find an address. In Arlington, the buses are also numbered: 41, 51, 52, 53 (A?), 61, 62, 67, 74, 75, 82. But what do these numbers mean? Like the streets, can I use the numbers to help me find the bus or figure out where it's going or derive any useful information at all? No. As a rider they are entirely meaningless. I asked around in Arlington County Commuter Services and virtually no one knew the "system" behind the numbers. Eventually I found one person who had a clue about the system but readily admitted that it was, in essence, arbitrarily contrived.

Metro is even worse. Not only are the numbers essentially arbitrary (I'm sure there's a system, but if it's opaque to the rider then it is no better than no system). Take the #2 bus. It comes in 6 varieties (2A, 2B, 2C, 2G, 2W. . . and 2T on a separate timetable). The 2W and the 2T are so different from the other 2's, they need a separate map and timetables. Why even call them # 2's? I'll bet you there isn't a single rider who can explain why those buses are called #2, why there are 6 different ones, why those particular letters of the alphabet are used, and why the 2W and 2T are somehow paired with the other 2's. The numbering is worthless to the rider (who is, after all, the customer). So here's a system that not only provides zero useful information but actually provides the disservice of confusing customers.
Imagine the streets were numbered like this: 8th, 11th, 5th, 1st, 4th, 9th, 2nd, 7th, 6th, 10th, 3rd, 12th. The numbers are meaningless (actually there's a "system"--revealed at the end of this post). If you tell someone you live on 5th Street you still have to describe where it is: "I live on 5th, which is between 11th and 1st." The numbers have become meaningless. Actually, this "system" is even worse than that; it creates more confusion than purpose. Better that the streets were given names like colors or trees.
"But how do you tell the buses apart?" Yes, of course they need to be identified. Everything needs some sort of identification: streets, animals, our friends, schools, devices, food. They all have names, and those names evoke meaning. Imagine if all the food in the grocery were just numbers. "Be sure to pick up some 22, 135, 16 and--oh yes--311 on your way home, honey." I find it hard enough to remember bread, milk and artichoke hearts, and--oh yes--toilet paper!
Eventually I would learn that 22 is milk, but it's so much harder. Our brains are not wired to apply numbers in an arbitrary way like that. We don't remember our friends by their phone numbers.
From early childhood we are taught that numbers are most usefully used as ordinal or cardinal identifiers--they help us put things in order or quantify them. But on buses they serve neither purpose (these are called "nominal" numbers), and so we have to deliberately undo a lifetime of learning and try to understand the number on the bus as nothing more than an abstraction that equates to a name. Better the bus be called the "phor" than 4; it would actually be easier. In fact, the metrorail lines being identified with colors is easier to remember than if they were numbered. And, although the colors are also essentially arbitrary, it is easier for the brain to bring meaning to them.

"But transit systems have been using numbers for generations, and it's been working." Thanks, Dilbert. Just because something has been done for decades doesn't make it good or leave no room for improvement. In fact, just the opposite: often it's the things that we assume ought to be a certain way are the things that should be questioned the hardest. Also, how do we know it's been working if we haven't tried something different to compare it with?
Boulder Colorado is one place I know of that has thrown out the number system (at least partially). Many of the buses have names: Hop, Skip, Stampede, Bolt, Dash, etc. The more complete names are things like "Skip Along Broadway" and "Dash down South Boulder Road." Now that's useful info to a customer. Personally I like the Jump (also called the Short Jump) and the Long Jump (which is the extension of the Jump--now there's a name that really works).
If the goal of transit is to help people get around better and more easily, that goal needs to consider everything: ease of use, cost, convenience, etc.. The names of the buses are a key piece of information critical to people using the system. Is what Boulder's done the best system? I don't know, but it's a lot better than everywhere else. In any event, the only system I can think of that would be worse than the arbitrary number system in common use is a system in which the buses have no identification at all.
====================
(originally posted November 10, 2007 on CommuterPage blog)
(Answer: the streets are listed in alphabetical order)

Squeezing More out of Metro with Operational Improvements

Metro ridership has been steadily rising for years. The Orange Line in Arlington and Fairfax, the "Orange Crush," has the worst crowding. In 1994, when I lived near Court House Metro, I could get a seat on my commute into DC. By 2000, I would only occasionally get a seat living two stops further out at Virginia Square. Now East Falls Church is my closest station, and most mornings it's unlikely I'll get a seat. There is no way someone at Court House ever gets one.

Metro has added some 8-car trains, which help. There is a fairly simple operational change that I believe can add capacity at no extra cost. Philadelphia has used this since 1956, and call it A and B trains. New York City called it "skip-stop" until they recently ended the practice.

Here's how it works. During rush hour each train is either an "A" train or a "B" train. Each train skips some stops. "A" trains start at Vienna and skip Dunn Loring and Virginia Square. "B" trains also start at Vienna and skip East Falls Church and Clarendon. (Skipped stops should be the least used stops, and should come in pairs, to balance ridership between the A and B trains).

The trip from Vienna to, say, Farragut West becomes 24 minutes instead of 27, a 10% time savings. The entire trip from Vienna to New Carrollton shrinks from 57 minutes to 54, saving 5%.

That time savings could allow Metro to save cars and run longer trains. There are more than 20 trains operating on the Orange Line during rush hour. Freeing up one train will allow 3 6-car trains to be extended to 8-car trains, thereby increasing the capacity.

The shorter runs could also allow more trains. Right now, Metro can't fit more actual trains through Rosslyn, but one day that might change if they send Blue Line trains up the Yellow Line, change signal technology, or build new river crossings. If it does, or if they try this on a different, less constrained line, Metro could run the same number of physical trains more frequently. Instead of 360 second headways, for example, they can reduce to 342 second headways, increasing the capacity of a line by 5%.

How does this affect passengers? Most will benefit, but some will be inconvenienced. Here are outcomes for commuters to DC from Virginia stations:

• Vienna to DC: 2 stations reduced ride time (BIG WIN!)
  
• Dunn Loring to DC: Increased waiting time and 2 stations reduced ride time (about a wash)
  
• West Falls Church to DC: Either 1 or 2 stations reduced ride time (WIN!)
  
• East Falls Church to DC: Increased wait and 1 station reduced ride time (slightly negative)
  
• Ballston to DC: 1 station reduced ride time (WIN!)
  
• VA Square to DC: Increased average wait and 1 station reduced ride time (slightly negative)
  
• Clarendon to DC: Increased average wait; no time savings (lose)
  
• Court House/Rosslyn to DC: No change

A small minority of riders traveling to one of the skipped stations also experience increased times waiting for their train. In a few cases, they may actually have to change trains if both their boarding and deboarding stations are served by different trains.

Vienna, WFC and Ballston riders all benefit, Clarendon riders lose about 3 minutes average waiting time, and Dunn Loring and EFC riders may be slightly negative or about even. The three benefiting stations represent roughly two-thirds of the ridership of these 7 stations, while Clarendon represents about 8% of the ridership. Therefore, two-thirds of the riders benefit directly. And everyone benefits from reduced crowding on the trains because of the increased capacity afforded by longer trains.

Plus, this same system could work on the other lines, too, including ones not at maximum capacity. It's probably easiest on the Red Line, since it does not have to merge with another line like the Orange does with the Blue and the Yellow does with Green.

They've been doing it for more than 50 years in Philadelphia. When I lived there (near a "B" station), it was just considered normal operating procedure. If it can work there, why not here?

==========================

(This post is cross-posted at Greater Greater Washington, along with almost 50 comments.)

Facing the Music

About a year ago I posted a link to an article written by Gene Weingarten of the Washington Post in which he convinced world-famous violinist, Joshua Bell, to play outside a subway stop and watch what happened. He won a Pulitzer Prize for the story.

Well, this Sunday's Washington Post magazine had a short follow-up to that original story, which is quite intriguing--particularly if you remember the original.

Enjoy!

How Long Does It Take? Really?

Have you ever noticed how people "adjust" how long it takes them to travel? When you ask people about their commutes, they will either undershoot to rationalize their particular choice or overshoot to make a horror story.

To whit: my wife recently changed her workplace to Silver Spring 3 days/week. We live in Arlington a 17-minute walk from East Falls Church metro. She drives to work, primarily because the walk-metro-shuttle-to-her-work trip would take almost an hour and a half each way. I encourage her to Metro, but she's justifiably resistant.

Recently our neighbor was over for tea and inquired about her new work. My wife said, ". . and it only takes 25-35 minutes to get there." After choking briefly on my tea, I challenged her on this.
Tuesday mornings at 7:30 my daughter has her piano lesson, and my wife has left it to me to get her there (without a car, because she has it). "Why is it then that you have to leave by 7:15 on Tuesdays to get to work by 8:00? You could drop [daughter] off at 7:25 and still make it to work in 25-35 minutes, right?" I asked.
Suddenly her story changed to 40-45 minutes. . .just to be safe.

Not to pick on my wife, but I hear this all the time. My experience is that drivers time themselves from the clock in their car, so they only count the time from when they pull away until they park. People who ride the bus or bike or metrorail count from when they leave home until they either enter their workplace or get to their desks.

I'm not really sure where I'm going with this. Just an observation.

HOT Lanes Won't Siphon Off Riders

I noticed an article in the Fredericksburg Free Lance-Star reporting on a VDOT report about future commuting behavior along I-95 when HOT lanes are completed. I was unable to find the report results on VDOT's site, so I have to depend on the Free Lance-Star for what details I could glean.

The gist of the survey results is that the vast majority of vanpoolers, carpoolers, sluggers and train riders will continue to use their same mode to commute along the I-95/I-395 corridor after HOT lanes are built and implemented. In fact, the type of commuter most likely to change their behavior is the SOV driver, presumably to another, more congestion- and environment-friendly mode. For people who are concerned that the construction of HOT lanes would siphon a lot of riders out of HOV modes into their SOV vehicles, this is a heartening result.

The survey polled 3,289 commuters. Here's how many said they would will stick with their current method of commuting:

95 percent of vanpoolers
91 percent of bus riders
86 percent of train riders
82 percent slugs
81 percent of carpoolers
53 percent of solo drivers

Mars and Venus (Earth Loses)

I'm often amazed at how really smart people can't seem to put 2&2 together. This last Sunday's Washington Post had a long, in-depth article about the BRAC relocations. It included a great graphic shown here (click on it for a larger view). At the same time, virtually every day brings additional news about the need to take immediate and significant actions to battle global warming. Most of the nations of the world met in Bali recently for the yearly Conference of the Parties that guides the international dialog on climate change. Europe, for one, was pushing for 25-40% reductions in greenhouse gases by 2020 (the U. S. 'succeeded' in watering down much of the language).

If you look at the map you will see that the jobs are primarily moving away from places where transportation options are more plentiful to places where they are not. There is no question that the number of miles driven in the DC area will go up with this realignment. The article points out the need for hundreds of millions of dollars of additional transportation infrastructure just to accommodate the additional traffic. All of it is more roads. Does the BRAC commission read the news about global warming? Are they living on another planet?

What you might hear from them is, "Our job was to blah...blah...blah. Once we met those objectives, then global warming might be a secondary issue, but it wasn't our job." I once attended a presentation in which the presenter posed the question, "Who is in charge of climate change?" The answer is, well, no one really. Which means that we all are. Including the BRAC.

It's probably too late to change the relocations, but it's not too late to make opportunities out of them. Knowing that all these jobs are moving, what changes can be made that will reduce the traffic and greenhouse gases at the same time? The most obvious is to look at which jobs can be done without needing the people to actually be there. Having workers telecommute 2-3 days per week (or full time) would reduce driving and congestion tremendously and also reduce terrorism risk by spreading out the workforce. Comprehensive additional planning should take place immediately, including starting work on extending rail (VRE and Metrorail to the Proving Ground and Fort Belvoir and MARC direct to Fort Meade). Other land use planning needs to be much smarter, both on the bases themselves and in areas that will attract workers. The forts themselves should make themselves examples of places that people can get around without needing a personal car. That way those who come by the train or bus or carpool can get where they want and need to be. One way might be to put in PRT (personal rapid transit), which I'll cover another day.

In any case, it would be great to see the Defense Department take the lead on being forward thinking and progressive on how to achieve their realignment goals without exacerbating global warming.

More DC Area Commuter Stories

The new Sunday column in the Washington Post on page B2, Commuter, today featured two real people and their actual commutes. I could relate to the first one almost exactly. He lives in Arlington near the East Falls Church Metro (I live in Arlington near the East Falls Church metro) and he works in Reston off Sunrise Valley Drive (I worked in a building about 100 yards from his for about 6-7 months in 2006). Pretty much identical commutes. He drives every day. I drove 2-3 times total in the time I worked there.

As in my post from earlier this month about why people don't use transit more, it appears that this commuter has not been able to get good information. He claims in the article that it would require 3 bus transfers to get from his house to his work. Not true--there is an express bus from the West Falls Church metro that goes directly to his office. It would take a bit longer than his drive (at least in the morning), but certainly a viable option if he's trying to avoid traffic coming home. I'm not sure how he compiled the information about the bus, but clearly he was unable to find good information easily.

However, his better option yet is to bike. The ride is virtually entirely on the W&OD trail. From the article:
"What would you rather be doing?
'I'd rather spend time with my friends, go out into nature, do some biking.' "

Why not do both at the same time? That was my way to and from work for the half-a-year I had the same commute. Biking and out in nature. I saw deer, fox, turtles, bats, groundhogs, birds, joggers, other bikers, and other flora and fauna. Sounds like the perfect solution for him. He's spending $50/week on gas (presumably not all just for his commute, though). He could certainly save some of that by leaving the car at home.

The other commuter lives in Waldorf and works in DC. I don't have any similar experiences to relate for him.

Perhaps most telling about why our traffic is so intractable was their replies to this question: "If you could change one thing about your commute, what would it be?"
Both of them replied that they wished everyone else would get off the road or use transit. This seems like a common sentiment (see my most recent post). . . if all those other people would just quit driving, then it would be so much better for me. Rokas Reipa

Bike to Work Day Success Story

This morning at the end of my bi-weekly commute to Silver Spring, another sweaty cyclist was sitting in front of my building, clearly having just finished his commute, too. We introduced ourselves and he asked where I bike from. I told him Arlington. He raised his eyebrows and said, "Me, too." (He is now the 5th person I have met in the last three months who bike commutes between Arlington and Montgomery County.) We swapped stories.

His story was that he decided to try out the commute for Bike to Work Day this last May. That convinced him that it was doable on a daily basis. He went out and bought a bike to meet his commuting needs, and now he's an every day commuter.

This was exciting for me to hear. Having participated in BTWD for a long time, I had the sense that everyone was already a bike commuter. It's nice to see--at least for this one commuter--that it was a catalyst for him to become a regular.

A Little Culture With Your Commute

If you didn't happen to see this article in last Sunday's Washington Post magazine, it is quite a compelling read. The set-up: world-acclaimed violinist, Joshua Bell, disguised himself as a busker at a Metrorail station during the morning commute and played for commuters emerging from the station. If you find yourself intrigued by the article, you can also check out the online discussion held Monday.

I suppose you can put a CD of Josh (or any other artist) on your car stereo if you're driving to work, but I doubt you could hire him to play in your back seat.

My Commuter Mug

This post is a little off the beaten path - it's more a savvy-consumer post. I generally take my insulated travel coffee mug with me, because I'm a tree hugger and I don't care to waste the paper cup (also, being insulated, it keeps my coffee warm longer). Interestingly, however, about 30% of the time I am given a significant discount (Starbucks sometimes just charges me 50 cents). Today, in fact, I wasn't charged at all; the server just said, "Nevermind, free refills." I'm not angling for the discount--I'm happy to pay for my coffee--but if it's offered, why not?

I think I've probably unintentionally saved at least $20 over the last six months just by having my travel mug with me. Try it yourself. Even if you're a paper-wasting environment hater, it might just save you a few bucks.