Pasteurs' work focused on disease prevention and a huge part of that is having clean, drinkable water. Unfortunately, many people around the world still don't have access to save water and the US National Academy of Engineering has made increasing that access one of its grand challenges of the 21st century. Any solution to a problem this big will require putting together lots of different engineering techniques. In practice, this requires making lots of choices and engineering design can help you make the right ones.

Only 2.5% of water on Earth is drinkable – the rest is mostly ocean water, which is much too salty. To access that water, you need to take out the salt, a process called desalination. We saw one way to do this when looking at mass separation techniques like Reverse Osmosis, sometimes called RO.

On a large enough scale, RO can turn millions of liters of seawater into freshwater every day. In fact, you could set up a desalination plant to do just that on an industrial scale. But scaling up from a simple idea into a fully-fledged plant requires some serious design work!

In industrial engineering, processes can get complicated fast. As usual, you have to keep track of the heat, mass transfer, and fluid mechanics for whatever you’re processing and how it reacts with materials in your equipment.

What’s more, every part will depend on another. So it’s not enough to string some techniques together to achieve your goal. Instead, you have to consider the process as a whole and the entire flow from start to finish. That means thinking about how to get water to where it needs to go, how to deal with waste, how to supply energy to the facility, and even how your plant is laid out. And of course, you’ll want to make everything as efficient as possible!

It might seem overwhelming to try and tackle all of this at once, which is why engineers use design work to make sense of a complex process. And usually, there’s a whole team of engineers involved!

The goal is to map out a problem and find ways to address its challenges systematically. You also want to ensure the whole system makes sense, and then test it to check that it does what it’s designed to do. The first stage is considering the main outcome you want to achieve and what techniques you’ll need to do it.

The central process of a desalination plant is using reverse osmosis on seawater. For example, by pushing the seawater through a membrane, water coming out the other side is more drinkable, while the salt gets left behind. That sounds simple enough, but there are actually a lot of steps hidden inside that idea.

First, you’ll need a way to get seawater out of the ocean and into your machine. You’ll also need a pump to drive the RO, a way to power it, and some way to get rid of the excess saltwater. Introducing specific equipment to address the needs of your process and then linking them together is called design synthesis.

At this point, you’ll also have to start thinking about where you can obtain the equipment, or the materials and labor you’ll need to build it. After talking with other engineers on your team, you might discover that there’s a supplier who makes water pumps that are already well suited for what you need. But that’s no good if they’re too expensive!

Even in the early stages, it’s best to start getting an idea of how much things are going to cost. As you add more pieces of hypothetical equipment in the design synthesis stage, it’s useful to keep a running total. Spiraling costs might require you to redesign or even abandon the project altogether!

That would be a shame, but it’s better to find out early on, rather than halfway through construction. We can assume that that’s not the case for the pumps and water tanks you’ve added to your design so far. Let’s also assume the materials and labour costs for building a custom RO unit also fall within the budget.

Finally, maybe you decide that the simplest way to dispose of the excess saltwater is simply to redirect it back to the sea through a pipe. With the basics of your design in place, you’re ready to start improving it. Engineers don’t jump from design to construction right away.

First, you’ll want to predict as much as you can about the performance of your plant to try and iron out the kinks. In fact, a big part of designing is redesigning based on what you find in this step You can get a pretty good idea of how certain parts will behave with some back-of-the-envelope calculations. As we saw in thermodynamics, simple equations like the maximum efficiency of a heat engine can give you an idea about the best possible performance you can expect.

But the details usually take a bit more work to figure out. If you wanted to estimate the efficiency of the RO unit you’ve designed, you could run a computer simulation to model its performance with different concentrations of salt. Computer modelling is often a cheaper way of investigating how complex pieces of machinery will behave, rather than building them outright.

Any problems you find can help you go back and improve your design. But Murphy’s Law rears its head here, too. Even the best modelling won’t capture all the real world behavior of your process.

Which is why, once you’re fairly certain about your design, it’s time to move to the bench scale. A bench scale model is a simplified, physical version of the process you want to put into practice – small enough to fit on a lab bench. Rather than processing thousands or millions of liters of water, your bench scale desalination unit will probably only handle something like 10 liters at a time.

Replicating your design with smaller, cheaper components lets you test the process from start to finish at low cost. At this stage, you might discover a critical flaw – like your pump bringing up sand along with the water! Maybe that sand clogs the intake, so water has trouble making it through the membrane.

Bench scale setups help identify problems like these where possible. To handle the sand problem, you could add an initial stage of filtering before the water reaches the desalinator.

At this stage, environmental considerations are also something you can’t afford to overlook. Industrial processes often produce waste, and it’s important to dispose of that in a way that’s safe and responsible. To avoid a pile of sand building up and blocking the output pipe, you need to arrange for it to be collected and transported elsewhere for agriculture or construction. After adding a sand filter and making sure it works on the bench scale, it’s time to jump up another level.

You’re still not ready to build a full desalination plant – but you’re getting closer! Just because a process works on a small scale doesn’t mean it will replicate perfectly as things get bigger. Consider the efficiency of your RO unit.

It might turn out that the surface-area-to-volume ratio of the membrane determines the system’s efficiency. Unfortunately, that quantity will change as you enlarge the membrane. Engineers call this situation non-linear scaling, and it’s a big challenge when designing an industrial process.

To account for issues like this, you’ll want to move on to pilot scale. The pilot scale model is almost a replica of your final plant, but usually only made to handle 10-50% of its final capacity. Some of the equipment might be the same as what you eventually use, but you’ll want enough flexibility to make changes.

When testing your pilot scale system, for example, you might find that the pump driving RO requires much more energy than your bench-scale calculations suggested! And maybe it’s so much energy that unless you find a way to bring down energy costs, it won’t be affordable to run the plant.

At this point, you might realize that you can put a power generating turbine in the pipe that channels wastewater from the desalinator out to sea. That turbine will allow you to recapture some of the waste’s kinetic energy, reducing your plant’s electricity needs and keeping costs down. Once you’ve resolved all the issues on the pilot scale, you’re ready to think about building the real thing!

Of course, knowing every part of your design and knowing how to make every part of your design are two very different things. For instance, you might not be able to install the RO unit until you work out exactly how far the pipe pumping in the water needs to extend. In general, the order you build things might not be the same order as things happen in your process. Which means you’ll need a construction plan.

Hopefully, you’ve been taking note of these things when putting together your bench-scale and pilot-scale versions, which will inform how the final construction of the plant should be carried out. And once you know how it’s going to be made, you’re finally set to actually build the thing! But a good engineer’s job doesn’t end there.

Once your plant is up and running, new issues might arise that need to be resolved. Maybe the price of electricity spikes a few years down the line, which threatens to shut down production. But you find out that the cost of solar technology has decreased in the meantime, which allows you to add a solar farm to directly power the plant at a reduced cost.

After all that work, your plant will be capable of safely and affordably delivering drinkable water for everyone nearby. In practice, a project of this size would require dozens of engineers working together to tackle all the necessary parts of the design process. It’s a lot of effort, but being careful and methodical in your planning gives you the best odds for tackling the biggest problems.

In this episode, we saw how design synthesis helps you put together the components of a process and decide what techniques are needed to solve your problem. We saw the need to test things on a smaller scale before ramping up to full production. Finally we saw how to continually incorporate feedback from design flaws to improve your designs.

Come back next week to learn what it takes for you – yes, you – to become a real-life engineer! Crash Course Engineering is produced in association with PBS Digital Studios. Keep learning with us and check out Physics Girl, where Dianna Cowern explains the physics behind puzzling phenomenon and everyday mysteries.

These are physics videos for every atom and eve. Subscribe at the link in the description. Crash Course is a Complexly production and this episode was filmed in the Doctor Cheryl C. Kinney Studio with the help of these wonderful people. And our amazing graphics team is Thought Cafe.