In 2015, an unusual incident happened on the  construction site for a sewage lift station  

in British Columbia, Canada. WorksafeBC, the  provincial health and safety agency, posted a  

summary of the event on YouTube. A steel caisson  had been installed to hold back soil while the  

lift station could be constructed. One worker on  the site was suddenly pulled into a sinkhole when  

the bottom of the caisson blew out. The cause of  the incident was related to groundwater within the  

soils below the site. We don’t all have to live in  fear of the ground opening up below our feet, but  

engineers who design subsurface structures do have  to consider the impact that groundwater can have.  

The solutions to subsurface problems are almost  always hidden from public view, so you might never  

even know they’re there. This video is intended  to shed some light on those invisible solutions  

(including what could have been done to prevent  that incident in BC). I’m Grady and this is  

Practical Engineering. In today's episode, we’re  talking about how groundwater affects structures. 

This video is sponsored by HelloFresh,  America’s Number 1 meal kit. More on that later. 

Groundwater has always been a little mysterious  to humanity since it can’t easily be observed.  

It also behaves much differently than surface  waters like rivers and oceans, sometimes  

defying expectations, as I’ve shown in a few of my  previous videos. One of the most important places  

where groundwater shows up in civil engineering  is at a dam. That’s because groundwater flows  

from high pressure to low pressure, and a  dam, at its simplest, is just a structure  

that divides those two conditions. And what do  you know, I’ve got an acrylic box in my garage  

full of sand to show these concepts in real life. You can imagine this soil sits below the base of  

a dam, and I can adjust the water levels on either  side of the structure to simulate how groundwater  

will flow. Blue dye placed in the sand helps show  the direction and speed of water movement below  

the surface. A higher level on the upstream side  creates pressure, driving water in the subsurface  

below the dam to the opposite end of the model.  I’ll be the first to say it: this is not the most  

mind-blowing revelation. You probably could have  predicted it without the fancy model. But to a  

civil engineer, this is not an inconsequential  phenomenon, and for a couple of reasons. 

First, water seeping below a dam  can erode soil particles away,  

a phenomenon called piping. Obviously, you don’t  want part of your structure’s foundation to be  

stolen from underneath it, and piping can create  a positive feedback loop where failure progresses  

rapidly. I have a whole video on piping that  you can check out after this one. The second  

negative effect of groundwater is less obvious.  In fact, until around the 1920s, dam engineers  

didn’t even take it into account (leading to  the demise of many early structures in history). 

The engineering of a dam is largely an  exercise in resisting hydrostatic pressure.  

Water in the reservoir applies an enormous force  to the upstream face of a dam, and if not designed  

properly, that force can cause the dam to slide  downstream or overturn. The hydrostatic force is  

actually pretty simple to approximate. Pressure  in a fluid increases with depth, so you get a  

triangular distributed load. Once you know that  load, you can design a structure to resist it,  

and there are a lot of ways to do that. One of  the most common types of dam just uses its own  

weight for stability. Gravity dams are designed  to be heavy enough that hydrostatic forces  

can’t slide them backwards or turn them over.  But, to the dismay of those early engineers,  

pressure from the reservoir is not  the only destabilizing force on a dam. 

Take a look at this pipe I’ve included in the  model that shows the water level between the two  

boundaries. If the base of a structure was below  the water level shown here, the groundwater would  

be applying pressure to the bottom, counteracting  its weight. We call this uplift pressure. Remember  

that the only reason gravity dams stay put is  because of their weight, so you can see how having  

an unanticipated force effectively subtracting  some of that weight would be a bad thing. Many  

concrete gravity dams have failed because  this uplift force was neglected by engineers,  

including the St. Francis Dam in California that  killed more than 400 people when it collapsed in  

1928. Many consider this to be the worst American  civil engineering disaster of the 20th century. 

Unlike the hydrostatic force of a reservoir,  uplift pressure from groundwater is a much more  

complicated force to characterize. It exists  in the interface between the structure and  

its foundation, in the cracks and pores of the  underlying soil, and even within the joints of  

the concrete structure itself. The flow of  groundwater is affected by soil properties,  

the geometry of the dam, the water  levels upstream and downstream,  

and even the subsurface features. How these  factors affect the uplift pressure can be pretty  

challenging to predict. But engineers do have  to predict it. After all, we can’t build a dam,  

measure the actual uplift force, and add weight  if necessary. It’s gotta work the first time. 

One way to characterize groundwater flow around  structures is the flow net. This is a graphical  

tool used by engineers to estimate the volume  and pressure of seepage in the subsurface.  

In simple terms, you divide the flow area into  a curvilinear grid, where one axis represents  

pressure and the other represents flow. If this  looks familiar, you might notice that a flow  

net is essentially a 2D solution to the Laplace  equation, which also applies to other areas of  

physics including heat flow and magnetic fields.  Developing flow nets is almost an art as much as  

a science, so it’s probably a good thing that  groundwater problems are mostly solved using  

software these days. But, we can still use flow  nets to demonstrate a few of the ways engineers  

combat this nefarious uplift force on gravity  dams. And one common idea is a cutoff wall. 

If water flowing below a dam causes so many  problems, why not just create a vertical wall  

to cut it off? We do it all the time. But, how  deep does it need to be? Some dams might have a  

convenient geological layer into which a cutoff  can be terminated, creating an impenetrable  

envelope to keep seepage out. But, many don’t.  Cutoff walls can still reduce the volume of  

flow and the pressure, even if seepage can still  make its way underneath. Let’s take a look at the  

model to see why. I’ve added a vertical wall  of acrylic below the upstream face of my dam,  

and we’ll see how it affects the flow. The  groundwater flow lines adjust to go under the  

wall and back up to the other side of the model.  If you look closely you’ll see a slight decrease  

in the uplift measurement pipe below the dam. The  only thing I changed between this model and the  

last one was adding the cutoff wall. So why would  the pressure decrease on the downstream side? 

The flow of groundwater is described with a  fairly simple formula known as Darcy’s law.  

Besides the permeability of the soil, the only  other factor controlling the speed water flows  

is the hydraulic gradient, which consists of the  difference in pressure over the length of a flow  

path. By adding a cutoff wall, I didn’t change  the difference in pressure between one side of the  

model and the other, but I did increase the length  of the flow path water had to take below the dam,  

reducing the hydraulic gradient. I can sketch  a flow net over the model to make this clearer.  

The black lines are equipotentials; they connect  areas of equal pressure. The blue lines show the  

directions of flow. Without a cutoff, the flow  paths are shorter, and thus the equipotential  

lines are closer together. With the cutoff wall,  the equipotential lines are spread out. That means  

both the volume of seepage and the uplift pressure  at the base of the structure have been reduced. 

Cutoff walls on dams have a long history of  use, and nearly all large gravity dams have  

at least some kind of cutoff. It can be as simple  as excavating a wide area of the dam’s foundation  

before starting on construction, and that’s  a popular choice because it gives engineers  

a chance to observe the subsurface conditions  and make sure there are no faults or problems  

before the dam gets built. Another option is to  excavate a deep trench and fill it with grout,  

concrete, or a slurry of impermeable clay. For  smaller or temporary structures, sheet piles can  

be driven into the subsurface to create a cutoff.  One final option is to inject high pressure grout  

to create an impenetrable curtain below the dam. The other way to deal with seepage and uplift  

pressure are drains. Drains installed below a dam  do two important jobs. First, they filter seepage  

using sand and gravel so that soil particles can’t  be piped out from the foundation. Second, they  

relieve uplift pressure by removing the water.  Let’s see how this works in my model. Upstream  

of my uplift monitor, I’ve added a hole through  the back of the model with a tube to drain seepage  

out. Instead of flowing all the way downstream,  now some of the seepage flows up to and through  

the drain, and you can see this in the streamlines  of dye flowing in the subsurface. Again, the  

effect is subtle, but the uplift pressure monitor  is showing a slight decrease in pressure compared  

to the original configuration. There is less  pressure on the base of the dam than there would  

be without the drain. Plotting a flow net over the  model, you can see why it behaves this way. The  

drain relieves the uplift on the base by creating  an area of low pressure below the dam. You can  

also note that the drain actually increases the  hydraulic gradient by shortening the flow paths,  

so there’s actually more seepage happening  than there would be without the drain. However,  

because the drains are installed with  filters to reduce the chance of piping,  

that additional seepage is often  worth the decrease in uplift pressure. 

Many concrete dams include a row of vertical  drains into the foundation, and some even use  

pumps to depress the groundwater level further,  minimizing the uplift. I can simulate this by  

lowering the downstream level as if a pump was  removing the water. Watch how the flow lines  

adjust when I make this change in the model. Like  drains, these relief wells create more seepage  

below a dam because of the greater difference  in pressure between the two sides, but they  

can significantly reduce the uplift pressure  and thus increase a structure’s stability.

I’ve been using dams as the main  example of managing groundwater flow,  

but lots of other structures have similar  issues. Retaining walls and temporary shoring  

have to contend with groundwater challenges,  including caissons, which are watertight chambers  

sunk into the earth to hold back soil during  construction. Remember the worker I mentioned  

in the intro? He was on a site near a caisson.  It’s typical to dewater a structure like this,  

meaning the water is pumped out, creating a  dry area for construction crews to work. Let’s  

take a look at how this works in the model. I’m  simulating the act of pumping water out of the  

caisson by draining out of the model at the bottom  of the structure. When a caisson is dewatered,  

it is essentially working like a dam, separating  an area of high pressure from low pressure within  

only a short distance between them. And, as  you know, distance matters when it comes to  

groundwater, because the shorter the flow paths,  the greater the hydraulic gradient, and thus the  

higher the volume and velocity of seepage. If you look closely, you can see the sand  

boiling up as the seepage exits the soil into the  bottom of the caisson. This elevated pressure in  

the subsurface and high velocity of flow means  that the soil particles themselves aren’t being  

strongly held together. All it takes is a little  agitation for the soil to liquefy and flow into  

the bottom of the caisson, creating a sinkhole  that can easily swallow anything at the surface.  

One way of mitigating this hazard is dewatering  the soil outside the caisson. Construction crews  

use well points, small evenly spaced wells  and pumps, to draw water out of the soil so it  

can’t seep to areas of lower pressure. Caissons  can also be driven deeper into the subsurface,  

creating a condition similar to a cutoff wall  on a dam. They can even go deep enough to reach  

an impermeable layer, creating a better seal that  prevents water from flowing in through the bottom. 

Thankfully for the worker in BC, his  colleagues were able to rescue him before  

he was consumed by the earth. Next time  you see a dam, retaining wall, caisson,  

or any other subsurface construction, there’s a  good chance that engineers have had to consider  

how groundwater will affect the stability.  Even though you’d never know they’re there,  

some combination of drains and cutoffs were  probably installed to keep the structure  

(and the people around it) safe and sound. 

Speaking of sounds, here is the peaceful  sound of a napping toddler while my wife  

and I prepare a home-cooked meal from  the sponsor of this video, Hello Fresh. 

“What do you call that?” “Hot raisin water” 

HelloFresh has kid-friendly recipes for  picky eaters but this one is just for  

us. The meals take about 30 minutes to  prepare which is perfect for nap time  

because it’s really hard to film yourself  preparing lunch while chasing a toddler at  

the same time, especially when the  photographer is 9 months pregnant.  

But… our little helper just couldn’t stand to be  left out of the fun, so he woke up early anyway. 

They also have different plans to meet your  various nutritional or dietary goals, like  

pescatarian, Fit and Wholesome, or vegetarian  which is what we usually get. It’s a perfect way  

to make a change or try something new. And this  is not us being mean, he really loves lemons. 

And here’s the most important thing: Hello  Fresh meals are delicious. We’ve done this  

for quite a while now, and there isn’t a single  recipe that we haven’t enjoyed. Actually we got  

invited to a party at the last minute, so we  ended up combining the portions of this meal  

and bringing it to share, and it was a huge hit. “This might be the best one you’ve made yet” 

Go try it yourself at HelloFresh.com, and if you  use code PRACTICAL16, you’ll get up to 16 free  

meals plus 3 surprise gifts. I wouldn’t recommend  it if I didn’t think you would love it. That’s  

HelloFresh.com and use my code PRACTICAL16. Thank  you for watching and let me know what you think.