The Dragonpipe (Mariner East 2) is to run entirely underground, with the exception of some pumps, valves, and vapor-control equipment. In rural most areas, that means digging a trench and laying the pipe four feet down.

But trenching is not practical in really built-up areas or under bodies of water like the many creeks in our area. I was curious about how Sunoco/ETP was planning to deal with that. It turns out that there is a technique called “horizontal directional drilling” (HDD) which is used for such situations.

Much of the Dragonpipe in Philadelphia’s suburban Delaware County and Chester County is to be built by drilling. This drilling will be done at a dozen or more sites in the two counties. It began at the first site, Judy Way in Aston, early in 2017.

I read up on HDD on the web, and this post summarizes what I’ve learned.

The HDD process proceeds in stages. First, a relatively small pilot hole is drilled. Then, the hole is enlarged by sending a larger drill or “auger” (or a series of them, in increasing sizes) through the hole. Finally, the pipe itself is pulled through. For the Dragonpipe, a final hole of around 30 inches will be needed for the larger of the two planned pipes (20 inches).

How do you get the drill to go where you want? When I first heard of HDD, my curiosity was aroused. How was it possible to stick a drill into the ground and have it emerge again, perhaps half a mile away, at a precise location? How could they know where the drill was, once it was underground? And how could they change its direction if it strayed off course?

The well drilling I had witnessed, growing up in the Midwest, was a simple process. Point the drill straight down, push the spinning drill pipe downwards until the end of it was near ground level, add another length of pipe to the drill string, and repeat. In the course of this process, drilling “mud” was pumped down the pipe to lubricate the drill. The mud flowed back up the hole and emerged around the drill pipe, bringing with it bits of rock and other material the drill had penetrated.

HDD, as I learned, is a refined version of that process. Two additional features are added, for changing direction and for monitoring the drill’s location underground.

Steering the drill. In order to be able to change the direction of drilling, the drill pipe cannot rotate as it does in traditional drilling. Instead, only the drill bit at the front end of the drill pipe is rotated. The drilling mud, which is constantly pumped down the pipe, is used as a source of power to turn the bit.

A short distance back from the bit, there is an “elbow” in the pipe. It is normally straight, but when bent it can make a slight adjustment in the direction that the drill is pointed. Electronics at the drill site control the bending of the elbow. The pipe can be rotated so that the elbow bend is in the desired direction: right, left, up, or down. Again, it is the drilling mud that serves as “hydraulic fluid” providing the force to change the angle of the elbow.

Monitoring the drill’s path. The other piece of the puzzle is knowing exactly where the drill is. There are several ways to monitor the drill’s location. The most common ones involve what amounts to radio communication between electronics in the drill string and sensors on the surface. In one version, the signals are monitored by workers doing a “walk-over” of the drill path with sensing equipment. In other cases, guide wires, sometimes with large electromagnets attached, are laid on the ground along the pipeline route. Generally, these electronic tools are sufficient to determine the drill’s location underground. Where communication between the drill and the surface is impossible, a gyroscope can be added to the drill string and used to gauge the drill’s approximate position, based on its last accurately-known location.

If the drill is found to be off course, the direction of drilling can be adjusted accordingly.

When drilling for a pipeline, sharp turns are not possible—even if the drill could be made to change direction abruptly, it would be impossible at a later step to pull the pipe around any but the most gradual curve. So constant, small adjustments to the pilot drill hole are essential. If pipeline route requires that pipe make a really sharp turn, that can only be accomplished by trenching at the surface and laying pipe in the trench, not by drilling.

After the pilot hole is completed and the drill emerges at the other end, the hole is enlarged in one or more stages. Then, the pipe is attached to the drill string at one end and pulled through from the other. As each section of pipe disappears into the hole, another is welded on.

How could anything go wrong? The description above represents how it works if everything goes smoothly. We have already learned that HDD doesn’t always go as planned. In May, drilling fluid from the Judy Way drill site started emerging in residential lawns and in the bed of Chester Creek. As I write this, it is still not clear exactly why this happened or how Sunoco/ETP is addressing the problem.

A future post will take a closer look at the problems that can arise from HDD. For now, the lesson to be taken is that HDD is not a perfect technology, and “pipeline” problems can start long before any pipe is in the ground.