The perfect storm

How Colorado's 1985 train wreck could happen again, this time with even more dire consequences

0

Rail traffic has increased in Boulder County in recent years largely due to the transportation of flammable crude oil, primarily from the Niobrara shale formation in Wyoming and here in Colorado. This is an important turn of events because the tracks that run through Boulder County have not been significantly updated since the last major railway accident on the Front Range — a 1985 head-on collision between two trains that ended in an impressive blaze that actually melted away a bridge on U.S. Highway 36.

Could a similar crash happen again, this time involving an oil train? How safe is a train line that at times cuts right through the middle of communities like Longmont and Fort Collins or runs right next to neighborhoods in Lafayette, Louisville and Boulder, where any type of derailment or explosion could prove extremely hazardous depending on the cargo?

So what caused the 1985 crash and what has been done to the tracks and the system used to direct rail traffic to prevent similar accidents from occurring today? The answer may surprise you. In the following pages, we offer a comprehensive look at the railroad tracks of Boulder County from 1985 to present, including an examination of the 1985 crash, a hard look at the current traffic and risks as well as an examination of the failed RTD deal that would have used these tracks for mass transit.

• • • •

Thirty years later, train crews along Colorado’s Front Range still refer to it as the day of the “perfect storm.” That is because it was a day when everything seemed to go wrong all at once.

Imagine yourself on Aug. 2, 1985 as a crew member of Burlington Northern Railroad mixed-freight train Extra 6311 West, leaving the Denver yard just before 7 p.m. and then stopping 4.5 miles later at Clear Creek near the Western Paving facility on West 64th Street. Once the train is fully stopped, you climb down from the cab, go into a booth and open the log book. The instructions in the train order you saw earlier were to stop at Clear Creek until gravel train Extra 6575 East has arrived and been switched onto the unloading tracks. You look down at the log book and see an entry that says Extra 6575 East came through at 5:45 p.m. on August 1. Somehow you have it in your mind that today is August 1, not August 2, and you don’t give it a second thought. You then climb back up into the cab, tell your fellow crew member that Extra 6575 East already cleared the main. Then the diesel engines begin their growling ascent uphill toward Westminster, Boulder and points north.

Seven miles later you look ahead and see the concrete and steel bridge where U.S. 36 crosses over the tracks on a lengthy embankment that completely obscures the rail line ahead. The rails curve sharply to the left as they pass under the bridge, making it impossible to see the headlight of an oncoming train. But you’re not paying much attention because you’ve traveled this route hundreds of times and you’re not expecting another train. You don’t even notice the approaching column of smoke that you might have seen above the highway ahead if only you had looked to your left.

Suddenly, there it is — the bright headlight of another train coming directly at you, less than 100 yards in front of you. Nine million pounds of steel and cargo are about to collide at a combined speed of 100 miles per hour. You rush to the door and leap and then everything goes blank.

This is how the accident happened, according to the National Transportation Safety Board. NTSB also reported that the explosion was heard 10 miles away; that a column of smoke was seen as far north as Fort Collins; that the U.S. 36 bridge was destroyed by a combination of the collision and heat from diesel fuel that burned; that no flammable materials were being hauled by either train other than the engine’s diesel fuel; and that five crew members died instantly.

It is easy to call this disaster a freak accident, the result of everything going wrong at the same time. After all, there was the misreading of the date and the coincidence of two trains meeting exactly at the one point on the entire line with least visibility.

This is all true. But to assume such chance happenings were the only causes of the accident is to miss a very essential point: that in 1985, the Burlington Northern line through the northern Front Range was basically a 19th century railroad, not a 20th century one.

Actually, its engineering wasn’t good even by 19th century standards. The portion of the line between Broomfield and Fort Collins was built in the 1870s by notorious “robber baron” Jay Gould, who sold it to the competition-wary Union Pacific Railroad at an inflated price in 1879. Union Pacific wisely replaced it within a few years by taking over a far better engineered route along the South Platte through Greeley and Brighton. And the portion of the line in Westminster, where the collision happened, was built in the mid-1880s to only slightly improved standards. By the early 1900s, the line, then known as the Colorado and Southern or C&S for short, had already developed the nickname “Crooked and Slow.”

Trains were still dispatched on this line in 1985 using the old method called “Timetable and Train Order,” a common practice of American railroads at that time for operations on unsignaled secondary lines. Burlington Northern locomotives had radios that could link them to the dispatcher, but the crews were not required to use them in all situations. Scheduled freights, of which there were several in each direction each day between Wyoming and Denver, simply headed down the line expecting all unscheduled trains to know to stay out of their way. Crews of “extra” freights, such as the two involved here, relied on train orders handed to them at the start of their journey telling them where to wait for which other trains.

Timetable and Train Order worked somewhat well when most trains on a line were scheduled, but in the 1980s the line was suddenly carrying many more unscheduled trains, including gravel trains like Extra 6575 East that shuttled daily between Western Paving and a quarry near Lyons. Extra trains were named according to the number on the side of the locomotive cab, so extra trains like 6575 that shuttled back and forth might appear in the log books multiple times each week, or even each day, easily leading to confusion by any employee who was mistaken about the date. As if that were not confusing enough, northbound trains were labeled “West” and southbound trains “East.” In Longmont and Fort Collins were line segments where “West” trains traveled eastbound and “East” trains westbound — another source of possible confusion.

And communication with the dispatcher, who in 1985 was stationed in McCook, Neb., was still done the old way too, using two wires stretched along a track-side pole line for all but 90 of the 490 miles between Wendover and McCook. In the summer of 1985 some of the poles were removed during construction of a new intermodal yard in Denver, and on the day of the crash, a temporary cable was run over and cut, leaving crews north of Denver temporarily without a link to the dispatcher. It is not known if this was a factor in the crash, as crews were not required to call the dispatcher, but it might still be added to the list of things that made Timetable and Train Order not a very safe way to run a railroad.

What has changed since 1985?

The good news is that Burlington Northern and its merged successor Burlington Northern Santa Fe (BNSF) have changed operating practices considerably in the intervening years. Within six months of the crash, Burlington Northern abandoned the Timetable and Train Order practice on all lines where it had been used. A new practice called “Track Warrant Control” was established. Basically each train now had to contact the dispatcher before and after every move and was given specific orders allowing it to occupy the track from location X to location Y and so forth. And the dispatchers had computer programs that prevented them from issuing overlapping track warrants.

At the same time, telephone booths were installed at most sidings and switches, using phone lines from the local phone company — phone booths that could be used in the event the company’s own communication lines were severed. These telephones were removed in the 1990s when dispatcher communication by radio became more dependable.

Since the late 1990s, BNSF has taken advantage of technological innovations as they came along, including cellular phones and satellite location devices. The dispatcher now has information on train location, speed and engine performance, and is promptly warned by computer technology of a potential problem.

BNSF has installed stronger, continuously welded rail (with extra-heavy rail on many curves), new ties and specialized clips that keep rails from forming “sun kinks” when hot temperatures cause them to expand.

BNSF has reduced at least a few of the sharper curves on the line. Near the Pearl Street Parkway crossing in Boulder, for example, two extremely sharp curves have been combined into one somewhat less sharp curve, although it is still sharper than many mainline curves.

What has not changed since 1985?

To begin, the Boulder County line, unlike most U.S. main lines, has never been upgraded to Centralized Traffic Control (CTC) or even Automatic Block Signaling (ABS). One of the things CTC and ABS do is send an electronic current down the rails. The wheels of a train create a short circuit causing track-side signals to turn red, thereby giving approaching trains a fail-safe warning to stop. Another important function is to turn signals red when a switch is not properly aligned for mainline traffic. And just as important is the ability of CTC and ABS to turn signals red when there is a broken rail. Not all broken rails are detected this way, but a significant number can be.

Then too, most of the engineering on the Boulder County line is still 19th century. Quite a few extremely sharp curves (six-degree or higher) still exist, one of them in a populated area of Longmont. Difficult-to-traverse grades of 1.5 percent and higher still exist too, often combined with six-degree curves — for example, near Arapahoe Avenue and 75th Street in Boulder, or on the hill just south of the Big Thompson River in Loveland. And “street running” still exists in Longmont and Fort Collins (i.e., the railroad runs down the middle of a city street for some distance). Drainage is poor in many spots along the right-of-way, leading to mud perennially working its way under the crossties and creating possible track instability.

Why have these things not changed in all these years? The answer is simple. This line was, until recently, of marginal value to its owner. Most freight on western railroads is east-west, not north-south. The freight that needs to come south from Wyoming to Colorado, other than coal and oil, consists largely of lumber and miscellaneous commodities, mostly from Washington and Oregon. Before the coal and oil booms, only rarely did this line see more than two through freight trains per day in each direction.

Once the coal boom started, and coal traffic became increasingly clogged on BNSF’s better but longer line through Northport, Sterling and Brush, the line through Boulder and Longmont was where overflow and detour trains were routed, at first mostly just empty coal trains going back north to the coal fields.

Now that we have an oil boom, and the coal boom has not significantly abated, BNSF is having congestion problems all over its system. First priority for additional sidings and second tracks is of necessity going to be in the Dakotas and in the Chicago area, where congestion is worst.

What can we be thankful for?

Paradoxically, perhaps, we might be thankful that the line through Boulder is not better engineered and that its deficiencies — the curves and grades, the absence of CTC or ABS signaling, and the dearth of full-length passing tracks — discourage BNSF from sending even more trains our way, especially heavy ones. If the line were better engineered, we would likely have six to 10 additional full-length loaded coal trains a day, plus who knows how many more oil trains. This is because BNSF’s other route, the one through Sterling, is 43 miles longer.

How much better is the route through Sterling? For one thing, it has hardly any curves sharper than five degrees, while the Boulder route has 56 of them. And the Sterling route has over 30 passing tracks longer than 7,000 feet, while the Boulder route has only five. Fortunately for us, perhaps, the cost of upgrading the Boulder line for heavy, full-length through trains, including coal and oil runs, is likely many times greater than that of improving the route through Sterling.

What dangers may await?

Oil trains (and molten sulfur trains as well) are, of course, a significant potential hazard, as local and national newspapers have been reporting for some time. But for an oil train disaster to occur there has to be a breakdown of some kind. Unlike the rail line through Lac-Megantic, Quebec, there is no siding on or on top of a steep grade, so a string of oil cars with brakes not properly set is not going to come careening down a hill and set a town on fire. Unlike the rail line near Lynchburg, Va., the track does not parallel a river with an embankment that can easily wash out under a train. And because of increased communication between locomotives and dispatchers, along with satellite location reporting, the chances of head-on collisions are far slimmer than they were 30 years ago (although certainly not out of the question).

What is far more likely along the Boulder line is a derailment caused by a broken rail and/or broken wheel. Even though the rail on this line is of average or better quality, and even though detectors that can pick up on damaged wheels are strategically placed, broken rails and wheels can still cause random accidents anywhere and at any time, possibly explosive ones if the derailed train carries crude oil or other flammable cargo. And loaded oil, molten sulfur and coal cars are six to seven times heavier than empty ones, thus many times more likely to cause broken rails or wheels.

Broken rails have been the leading cause of derailments throughout the U.S. and Canada for many years. News reports have told of many derailments — 10 major ones just in the month of March 2015, two of which were in Colorado (Hudson and LaSalle) — and every one of them involved trains with heavy loads (coal, crude oil or molten sulfur). Clearly railroads have pushed their car weights to the limit. All it takes for a rail to break is to have a couple of split or rotten crossties underneath it while being pounded by string after string of oil or coal cars weighing around 140 tons apiece.

This is not to say a disastrous derailment will occur tomorrow or next year or even in the next 10 years, with or without heavy trains hammering on the rails. But no one can guarantee such a tragedy will not happen.

What can Front Range cities do?

Unfortunately, local communities, or even states, cannot restrict what commodities railroads carry. Laws and legal precedents going back at least 100 years prevent this. So, then, how can cities like Boulder and Longmont keep dangerous cargoes from passing through?

For some years Denver Regional Transportation District (RTD) has been interested in using this line as a means to extend commuter rail to Boulder and possibly points north. If the line were to be purchased outright by RTD, that sale would theoretically put an end to through freight traffic on the line for good.

The argument has been raised that local freight from customers in Loveland, Longmont and Boulder would still need to travel via the line. However, Loveland, Lyons and Longmont traffic could easily be handed off to the Great Western Railway, which could transport it to Union Pacific’s Greeley line at junctions to the east, and local freight from Boulder could go the same route if new track were to be laid on some abandoned roadbeds north of Lafayette and/or north or east of Erie. A small amount of freight, including lumber cars from Lafayette, grain cars from a plant north of Longmont and loadings from future new customers might still need to be handled on the Boulder line, which is certainly doable.

But RTD’s negotiations with the BNSF Railway make for depressing reading by anyone hoping to see fewer freights and more “heavy rail” commuter cars. RTD missed its chance in 2005 when it backed out of a possible deal, claiming BNSF was asking for too much money, because by 2013 the price had gone up eightfold to $535 million. That figure was not actually to purchase the line outright, but simply to have trackage rights to run commuter trains on the same tracks during certain time slots. BNSF would have needed the money to install signals on the line and to double-track some segments, at least between Denver and Boulder or Longmont, for RTD’s benefit.

What has changed since those last negotiations is the appearance of more oil and molten sulfur trains on the line, fueling residents’ fears of a possible fiery crash in a residential neighborhood. Six hundred thousand people live within a few miles of the line as it twists its way through Westminster, Broomfield, Louisville, Boulder, Longmont, Loveland and Fort Collins (compared to only about 60,000 along the three-state Sterling route).

As already mentioned, BNSF would appear to value this line almost solely as an overflow and detour route. This is for the simple reason that upgrading it would likely cost far more than adding a similar amount of capacity to the Sterling route. Lengthening sidings and adding more track are one thing; getting rid of the curves and grades that give loaded coal and oil trains a hard time is quite another, involving construction of trestles, embankments and cuts, not to mention purchasing property. And what would it cost to bypass the “street running” in Longmont and Fort Collins?

Fortunately, the cost of upgrading the line between Denver, Boulder and Longmont for commuter trains is far less than upgrading the entire Denver-Boulder-Longmont-Cheyenne- Wendover line for more freight trains. The street running is to the north of downtown Longmont, so a commuter line from Denver to Longmont, if built, would stop short of it. And the tracks up Atwood Street could be torn up, or at the very least kept in place for a few short local freight train moves each week.

Curves and grades are not as big a deal for commuter coaches as they are for full-length coal and oil trains. Commuter coaches typically weigh less than 25 percent of what loaded coal or oil cars weigh and can easily handle 2 percent grades and seldom cause broken rails. Portions of the line would still need to be double-tracked and a few of the very sharpest curves reduced, but there would be little or no need for building trestles, embankments and cuts to bring commuter service to Boulder and Longmont.

BNSF is currently spending billions upgrading its trackage and adding more tracks and longer sidings, but almost all of it is in North Dakota, Wisconsin and Illinois. Eight locations in Colorado will get longer sidings and/or improved terminal facilities this year, but all eight are on the Sterling route, not the Boulder one. It would stand to reason that if BNSF continues upgrading the Sterling route, there may come a time when the Boulder line is even less valuable to the railroad than it is today.

If a detour route is still needed, BNSF should be able to route via trackage rights on Union Pacific through Greeley and Brighton — at a cost, of course. Union Pacific already has trackage rights on BNSF to bring coal to Boulder for the Valmont Power Station. BNSF could easily secure similar rights on Union Pacific and Great Western for any new oil trains that originate along the Front Range; those trains could travel east across branch lines to reach the Union Pacific and then be switched back to BNSF in Commerce City or Denver.

Currently, BNSF has nine crude oil loading sites in Wyoming and Colorado. Fortunately for us, seven of the nine are far enough north that their loads can be hauled via BNSF’s Sterling line to the east of us. The other two, Cheyenne, Wyo., and Windsor, Colo., are located in places without access to the Sterling line but with easy access to the Union Pacific. Trackage rights over Union Pacific would be a possible solution.

And a map of northern Colorado shows that Weld County, where most additional fracking is likely to take place, is very well covered by the Union Pacific and the Great Western, while BNSF is hardly a player.

What would happen if Boulder County were to allow fracking? It is not the location of the wells that would be a concern, but that of the loading sites. Very likely any new sites would be to the east, even if some of the wells were situated to the west. Oil is nearly always trucked from wells to loading sites.

A very efficient method of running trains is called “directional running” and is what is used on the Colorado Joint Line between Denver and Palmer Lake. In this case, trains of both railroads could go in one direction via BNSF’s Sterling route and in the other direction via Union Pacific’s Greeley route. The efficiency of both routes would be nearly doubled this way — without huge investment, except possibly on the portion of BNSF’s line between Denver and Brush, where trains would still need to be handled in two directions. Unfortunately, BNSF runs far more trains between Wyoming and Colorado than Union Pacific does, so it would not be an equal exchange and would cost BNSF money, although still possibly less money than increasing capacity on the Sterling line.

Perhaps it is time for all parties to sit down at the table once again — BNSF, RTD, the communities along the line and possibly even Union Pacific and Great Western — and look at all possible options and figure out what each party would need to contribute to get BNSF to run all its trains farther east. Who knows, BNSF might actually be more willing than before to dispense with its “Curvy and Slow” line, especially if selling it would give it quick cash to help gets its trains moved to better engineered and less populated routes to the east.

Clearly, the oil boom has changed the game, bringing with it new hazards, new challenges and, most importantly, new opportunities for negotiations — negotiations that might prove to be the only way to keep our rail line from having another one of those days when everything goes wrong.

Paul Gibb is a former English professor and lifelong student of railroading who at the time of the 1985 train crash was working as an electronic technician in the Signal Department of the Burlington Northern Railroad, assigned to the Denver Region. He has lived in the Boulder area for 30 years.