Monthly Archives: May 2015

Pressure & Flow Restriction

It was terrific that the last post was so well received. It would seem that from others’ experiences and some more of my own that a lower pressure/flow has the potential to improve extractions irrespective of the grinder. The suggestion being that lighter roast coffees are harder to extract from evenly as a pressurised pour over (aka espresso) and that a lower flow rate helps with this.

The dialogue on the topic of pressure has mostly suffered and been stifled by the conceptual differentiation of flow rate and pressure. I was motivated by this misconception and wanted to explore the language and theory in this post.

We can start by stating that flow rate applies only to water and pressure refers to only the pump. In an espresso machine, the pressure determines the flow rate of water, but this pressure is not the hydrostatic pressure of water.  The pressure in an espresso machine refers specifically to pump pressure: whatever technology the pump is based on. Flow rate and pressure are not the same thing when dealing with gases, but all we care about when talking about coffee extraction is the rate at which water is pushed through the coffee bed.  Before we can explore the physics, we must go on the journey water takes through the machine.

When we make an espresso, water is forced towards the group using some sort of pump.  This force comes from applying pressure to the water.  But this does not mean that the water is pressurised.  Where did the word ‘pressure’ come from in terms of espresso machines?  Pressure is an important measureable for gases and most espresso machines force water through the pipes using some form of pump.  This is very different to applying ‘pressure’ to water.

Now that the water is moving, it progresses through the group, followed by exiting the shower screen.  It then comes into contact with a resistant barrier (the packed coffee dose).  Instead of permeating through the coffee, water is continually pumped into an open head space above the coffee. Once all of this space has been used up and there is nowhere else to go the water is forced through the coffee.

The coffee and the holes in the basket are effectively the exit point for the water. They’re simply resistors, so the flow rate is defined simply by the rate in which the coffee travels through these resistors.  We use liquid water to make coffee, which makes the discussion of pressure somewhat easier. Unlike something like air that will readily compress significantly under pressure, water is almost completely uncompressible. For example, at 4km beneath the ocean’s surface water only compresses by 1.8% by volume. Hydraulic cranes work by filling metal sleeves with water that will not compress even when lifting very heavy objects.  We know that water expands when temperatures deviate from 4 ºC, but, in terms of an espresso machine, the density of water is only a function of temperature, not pressure, and this means that we can measure the flow rate in L/min.  Any object in the way of the flow of this water will slow its rate.

It gets tricky here, because in an open system (where the water can leave by for instance flowing out of the basket) means we must consider displacement. The water entering the basket displaces the water already occupying the basket. Given that the basket (including coffee) is the resistor; the rate of displacement defines the “pressure” in the basket. Note that this effective pressure is simply measured by flow rate, and hence if the pump pressure decreases the flow rate does too.  Not because the water itself is less pressurised, but because the force pushing the water is less.  When we increase pump pressure we increase the flow rate of water exiting the shower screen and vice versa for a lower pump pressure.

There is another way to lower the flow rate coming out of the group head called a flow restrictor.  A flow restrictor is basically a little pipe that the water passes through during its path to meet the coffee. It effectively funnels water into a narrower pipe, and like traffic on a motorway getting pushed into one lane, the same volume of traffic will now take longer to pass the same distance effectively slowing the traffic rate.

If we consider a machine that has a pump achieving 9 bar pressure pumping water through the group. We then install a narrower flow restrictor in the path. This slows the flow of water down at the point of the flow restrictor, which directly decreases the flow immediately after the flow restrictor.  Our  pump pressure remains high,  but the rate at which the coffee brews goes down.  We have effectively lowered the flow rate using a flow restrictor rather than just lowering the pump. There are mechanical issues to consider but in essence the flow restriction and pump pressure impact flow rate the same way through different modes.

It’s much easier to just adjust the pump rather than exchange a flow restrictor.

To help illustrate this problem, below is an image that show that as the flow restrictor becomes narrower, there is a greater difference between incident flow rate (i.e. incoming pump pressure) and exit flow rate (i.e. the flow of water after all resistors have acted on the water).  Furthermore, you can achieve exactly the same effective flow rate with any flow restrictor, Simply by adjusting the pump pressure to compensate for the flow restriction.

This first image displays how different in going flow rates create different out going flow rates with different flow restrictors

restrictor

This second image displays how flow rate going in can be altered to achieve the same outgoing flow rate when the flow restrictors are different.

pressure

It would appear that discussing pressure and flow restrictors over complicates the discussion. In terms of sharing brewing parameters it makes more sense to describe the amount of water that exits the group in a given time with no resistance. This way we implicitly understand the relationship between pump and flow restriction on a given machine.  It also removes the obvious variable of grind size, which directly determines how strong a resistor the coffee bed is.

Leave a comment

Filed under Uncategorized

15g baskets

We have moved over to using 15 gram baskets to brew all of our espresso in store.

I have explored in other posts how our recipes and practices have changed over time. How the value of ristretto style shots became questioned , especially if the goal is the fullest character in the cup. Then how the very definition of espresso has become more malleable and ranging. Longer shots, lungos and the like have gone from being seen as travesties to tentatively and logically considered, to then being fully recognised as positive ways to brew coffee.

Of course, brewing is never in isolation. It is linked to the coffee, grinders, water and roasting. Light roasting can work very well but it is also in danger of not fully developing the coffees character, which is reason in the first place to move away from darker roasting. But baked coffees start to pop up and under developed coffees become very common, both in the flavour form of light grassy sourness or as a brothy vegetal toastiness(where the coffees isn’t developed  but has been roaster darker to compensate).

It can sound obvious, but many roasters I speak too are looking to now roast the coffee as fully as possible before they get the taints associated with pushing it too far – ashy, toasty, heavy. This is a seemingly small but significant conceptual move away from roasting it as light as possible before it is too light. After all, its the roasting process that harnesses the coffees potential and produces flavourful by products. Roasting is a tricky subject and I am excited to explore it in depth ourselves shortly. It’s also great to have more openness about roasting theories within the community, much like the world of brewing has seen, especially as it is all intrinsically linked.

A full even extraction is part of this process of understanding. It’s a complex feedback loop. Roast the coffee to be more soluble and you can achieve shorter shots with higher extractions and vice versa, in that less soluble coffee will need more water to achieve a higher extraction. Seriously underdeveloped roasts will struggle to reach high extractions at all, regardless of the amount of water. Solubility is an interesting indicator of roast and it’s a piece of terminology that is becoming more and more used. Roasting coffees to the same solubility is a concept that I have seen more of. It is also curious to consider how the same solubility could be achieved with different flavour development. But like Extraction numbers in brewing, it has the potential to improve the data and the conversation.

There’s a lot I want to taste in a coffee, I am looking for espresso to have intense complexity accompanied with sweetness whilst also being clean. I also want it to have the body and tactile qualities that set espresso apart (I’m a big fan of lungos, but see them a s different drinks).

In store we have moved to brewing all shots on the EK43. Our preferred recipes often ended between 35 and 45 percent brew ratios. The only downside is that the shots were getting long. Simply put, I didn’t like the way they sat in the espresso cup, they were also often too much. We have In our store always presented espresso as a one size beverage  as opposed to doubles & singles.

Over the years we had dropped down from 20g baskets to 18g baskets. We then thought, hell, why not go even lower?

So we got hold of some 15g vst baskets. The plan was to simply scale down. Keep the same types of ratios and extractions but just have less of the same drink in the cup, effectively reaching a desired portion size.

My initial concern was that it wouldn’t be this simple. The diameter of the dose remains the same, but the depth decreases. This presents questions for both flow and temperature.

Is a shallower bed less or more likely to extract evenly? Also, with a higher percentage of the dose now being immediately accessible to the water, as well as less overall temperature decreasing energy from the dry coffee, will the overall temperature of extraction remain higher?

Our biggest challenges with the drop in basket size have all been to do with restriction in the basket and the ability to get consistently high and even extractions. VST acknowledge the fact that as the basket depth decreases and we use less coffee, the amount of resistance the water faces, changes. The pressure in the basket effectively decreases as the amount of coffee decreases. VST aim to counter this by changing the frequency and size of the exit holes. The goal being to achieve the same resistance basket to basket. Ideally, the same grind should result in the same ratio to coffee to water in the same shot time.

For us it didn’t work. This may be to do with the grind distribution or the roasting styles. With the coffee burrs we just couldn’t get the shot slow enough, and the extractions high enough. We then moved to the Turkish burrs. This did allow us to achieve slow enough flow rates. The extractions went up but they weren’t consistent. The crema was erratic. Finer again didn’t help matters. We were hitting a celling on extraction and it would appear that by going so fine we were actually lowering the extraction, with the bed creating pockets of clumped fines that hindered the flow of the water, in essence having a coffee puck full of channels. This was something I had heard of before but had not encountered so obviously.

Next we took the dose up to 16-17g to increase resistance/pressure so that we could grind a bit coarser and hopefully achieve more even extractions. We collapsed heavily to keep head space, evenly distribute the bed and worked a nutating tamp in to the mix. Results improved. We also dropped the temperature from 95 to 93 to compensate for the potential increase of overall brewing temp and this also slowed the flow down again. This was all with the Turkish burrs. I still preferred the results I could achieve with the coffee burrs in bigger baskets.

The real success came by taking the pressure down to 6 bar. I was thrilled to have someone suggest this to me. On the San Remo Opera this gives us a flow rate of 190g of water in 30 seconds (straight out of the group with no handle inserted). If the pressure and flow are relative to restriction then it makes sense to lower pressure as restriction is lowered.  This has allowed for use of the coffee burrs again. The results in the cup are top notch.

The exploration of this basket size, was as you can see not without its problems, but the successes are multiple – espresso beverages of our desired serving volumes, with full, even and flavoursome extractions, that saves us coffee and money. Happy days.

Comments Off on 15g baskets

Filed under Uncategorized