This week, Scott Rao made espresso with filter papers above and below the coffee bed yielding over 25%. What’s he playing at?

If you follow Scott Rao’s series of daily coffee tips on Instagram, you’ll have seen one recent post where he describes using paper filters cut to fit the espresso basket, placed above and below the puck. By doing this, he is able to reduce channeling, and push extractions to 25% and beyond, and get tasty espressos even at these high extraction percentages.

‘Not only does each filter seem to increase extraction, but the top filter seems to mitigate channeling — a lot,’ Scott writes. ‘This is a very big deal.’

While people have already been experimenting with Aeropress filters in espresso baskets for some years, it seems there’s something pretty interesting happening if it’s possible to achieve such high extractions without ‘over-extraction’ flavours. So what part of this approach is new, and why do paper filters have such a big effect on espresso? And perhaps most importantly, why don’t these shots taste over-extracted?

How Does This Work?

First – let’s look at what we already knew. If you’ve ever experimented with putting a paper filter inside your espresso basket, you might have been surprised to find the shots run considerably faster, even though you’re putting an extra layer in the way of the flow. During an espresso shot, coffee particles, especially fines, ‘migrate’ – moving with the flow of liquid towards the bottom of the puck. In a traditional espresso basket, these particles partially block the holes in the bottom, slowing the flow. With the paper filter at the base, the particles can’t pass through to reach and block the holes, so the flow is faster. The leap here is remembering that a faster flow allows you to grind finer, giving higher extraction. In espresso, any method that allows higher extraction is effectively giving you more even extraction, which leads to better flavour and more sweetness, and allows you to use less coffee to achieve the same strength.

What we at BH hadn’t seen before is the step of adding a paper filter above the puck as well, which further increases extraction. “As with many things in coffee, we know it works, but the “why” is mostly speculative,” Scott says. “I assume the filter on top helps the dispersion of water on top of the puck… and perhaps helps keep the top of the puck smoother and more intact.” This would be similar to the reason cold drip towers often use a paper filter on top of the coffee bed, to prevent the drips carving a channel through the bed.

What About Over-Extraction?

When combined with other methods to increase extraction, Scott is making espressos at even higher extractions – over 27% – and with that, approaching the limit of coffee’s solubility. Yet he claims that the shots taste better than ever, without the bitter and dry flavours we’re taught to associate with over-extraction. How is this possible?

The answer is all down to channelling. Those dry and bitter flavours are due to tannins and other large molecules, which can only dissolve in areas of extremely high localised extraction — in other words along the path of a channel. Over to Professor Steven Abbott to explain why this is:

to explain why this is:

“There’s a law of physics called the no slip boundary condition which means that the velocity of a liquid flowing past a surface is exactly zero at the surface itself. This means that the only way for molecules to escape from the surface is via diffusion, and for large molecules like the bitter tannins this is very slow – a good thing because we don’t want them. The only way for these molecules to escape is to get turbulent flow coming as close as possible to the no slip boundary – and the only way to do this is via the Forcheimer, rather than Darcy flows mentioned in the previous post. Channelling is a sort of Matthew effect – the more flow starts down one path, the more turbulent it gets and the more it flows, creating more turbulence and dragging out more tannins. So anything which reduces channelling will help reduce the bitter extraction.”

In other words, it’s not just over-extraction:  tannins and other large molecules need the turbulent flow created by channels to extract into the coffee. Because our tongues are so sensitive to bitterness and dryness, it doesn’t take much of these molecules to ruin an espresso. However, if we’re able to reduce channeling, then we can push extraction higher before these flavours start to become apparent. “‘Over-extraction’ is an overused term,’ Scott explains. ‘These days I never use the term except to describe localized over-extraction due to channeling.’

The SCAA recommended range of 18-22% was based on established equipment and techniques, and stood for many years. There’s been a steady trend however in recent years towards aiming for higher extractions — enabled by better equipment, better roasting, and better barista techniques. This began with the use of better grinders like the EK43, allowing extractions at perhaps 24% before over-extraction flavours became apparent, but has continued since as more techniques to promote even extraction have emerged.

“Mind you, not all extractions over 26% taste great,” Scott says. “I see this is a new frontier where step one is to raise the ceiling on the extraction level and step two is to begin optimizing coffee at these higher extractions.”

As to how to keep raising that ceiling? We at BH have been discussing an idea with Professor Abbott that could take this a step further: “A proposal for a drastic solution to the channelling problem is to have such a loose bed that there is nothing to channel, with the hot water free to just extract all the good stuff,” he explains. “But what about back pressure? First, why do we need it? Partly because without it the CO2 that explodes out of the beans can get in the way of the water (which is why we bloom open beds of coffee). And partly to get the crema that we all love. So let’s get the pressure via a different means – a filter. Not the usual “easy flow” filters we are used to. But the very fine filters used routinely in the bio-world for filtering out, say, bacteria. A look at the data sheets suggests that with a 9 bar pressure, the flows would deliver a shot in similar times to regular espresso. And by controlling the flow independently of the coffee, we are free to, for example, over-grind the coffee with plenty of fines. But won’t fines block the filter? We might call them fines, but to a bio-grade filter, these are huge boulders, the filters won’t even know they are there. That’s the theory. All we now need is to test it.”