01 May, 2012

h20 : part 3 of 3

Reverse Osmosis (RO)
Now that we have out minds wrapped around Ion-Exchange we can move onto the other major way mineral and/or dissolved solids are removed from water: Reverse Osmosis (RO). To best understand this process lets break down water into two components: solute and solvent. The solute is everything other than h2o, all of the dissolved (and particulate) solids. The solvent is the water in its pure form. A reverse osmosis system is made up of six main components: a pump, a taste and odor micron filter, system membrane(s), mineral cartridge (calcite), post scale inhibitor filter, and a storage tank. To understand this process best we'll focus on the membranes as they are the component that does the removing, or more accurately, the separation of the solute from the solvent. Reverse osmosis at its core is based on a pretty basic principle in chemistry that we touched on earlier. It is the tendency towards balance at the molecular level. In our previous example as it relates to ion-exchange we talked about positively and negatively charged ions. In the natural process of osmosis when two liquids that contain a different amount of solute (or TDS) are separated by a permeable membrane, there is a tendency for the lower solute liquid to move towards the higher solute liquid to achieve a balance of solute concentration. This tendency in our case is referred to as osmotic pressure. During reverse osmosis pressure is increased (via an external pump) to reverse this process and separate the high-solute liquid from the solvent. The high pressure is very important as only under the higher pressures can the chemical process separating the solvent occur in reverse. When we brew espresso we are brewing at about 9 bar of pressure. When you lose the pressure, you can no longer emulsify the oils from the coffee and you lose flavor, crema and really, you lose what we define as espresso. Now, RO is really nothing like that, but it does require high pressure to accomplish the goal, 17 bar to be specific. 250psi of pressure force the water into the membranes. The membrane is best pictured as the front page of a newspaper wrapped up very tightly: multiple layers in a circular/spiral pattern. The membranes are designed such that there is only one-way movement from the high-pressure/high solute side to the other side: low pressure/low solute. Under these circumstances the ions (mineral, sodium and otherwise) are removed and only the h2o molecules pass through the membrane. The high-solute water is directed or washed out of the membranes as waste water and the filtered water which is now almost or at 0 TDS comes out the other side of the membrane. Since the water exiting the membranes is so aggressive it is routed through a cartridge much like the bowl style calcite cartridge we talked about earlier. This 'remineralizes' the water and 'stabilizes' it for use. In some systems it is then run through a scale inhibiting filter before hitting the equipment. The inherent problem we are faced with, however, is that water goes into the membranes at a very high pressure but comes out at an extremely low pressure and flow rate. There are many sizes of systems and in general, the larger the membrane the higher the output flow-rate and pressure will be, but it can range from a drip/trickle to garden hose half turned on type flow-rate and pressure. From here on out I'll use flow rate and pressure as one in the same even through liquid science dorks would really criticize me, the normal people out there like me need to take it one concept at a time.  Since the output is so little we are faced with a problem. Our coffee brewers need at least 20psi to operate. And we need minimum flow rates and stable pressures to brew espresso too. With an RO system you'll see a small to rather large tank down-line from the rest of the system. About half or more of the tank is taken up by a large inflatable bladder. It is kind of like a bike tire. A bladder inside is pressurized to keep a certain amount of resistance against the tire material. Since water cannot be compressed it meets the resistance of the bladder inside the tank which compresses to then apply pressure to the water tank. With a large enough bladder and volume of water, stable pressures can be achieved. RO systems are very effective at taking care of water chemistry problems. However, the main concern on my end is two-fold. One, there is an amazing amount of water waste involved in most RO systems. Two, stability in dynamic water pressure and flow is pretty tough unless the system is very large in comparison to your needs.

Water waste is an inherent reality in the use of an RO system to filter your water. The discharge water that is highly charged with the solute is routed to a drain. Unfortunately the efficiency can be as low as 15%. For example, of the 100 liters of water that the system receives only up to 15 liters makes it through to the storage tank while 85 liters goes down the drain. Now, large commercial systems can achieve efficiencies higher than this, but we still see a lot of waste during RO. This is a problem. I'm sure someone could probably come up with a good use for the waste water and reroute it for storage. Or use it for the toilet flushing, garden watering or better, but I can't say I've ever seen anyone with an RO system doing that.

Stability in espresso brewing pressure has been a major focus in the specialty coffee industry. Dialing in multiple pump pressures, flow restrictors, and valves to give the barista ultimate control over brew pressure has been a realization (I posted here) in the espresso machine world. Honestly, the most stable brewing systems would be a result of the machine drawing filtered (RO or whatever) water from a water tank at 0 pressure (think trade-show set ups). In an RO system you are completely at the mercy of the holding tank and the pressure it can give you. If your usage out-paces the systems' ability to process water then you will quickly run into erratic pressures. This can be fixed with larger membranes and larger storage tanks, but is a real problem with many of the systems marketed to espresso professionals.

Some of my opinion towards RO systems has to do (unfairly) with never really seeing them work well consistently. Whether it is the user's fault for not changing membranes on a regular basis, the system being under-sized for the demand, or water not being properly pre-treated to make the system works properly, I haven't seen great performance. Based on my experience, even Starbucks has more problems with the water system (RO) then the rest of their coffee equipment combined. As users meet these issues, costs get pretty big. Storage tanks are made out of special materials to handle aggressive water, membranes are expensive and in larger systems, a pain to change. With all of the cost and maintenance involved to keep the system up and running it does offer nice water. But again, the user needs to make sure the PH and Chloride levels are okay for usage as well. Bonus: Most RO systems should use softened water instead of hard water. The membranes will last longer removing the sodium ions from soft water than the mineral ions from hard water. Also, a pre-filter that takes water down to .5micron clean will help the membranes last longer.

pH, Chlorides
Aside from the TDS and filtration tactics that we've already covered there are two other aspects that are very important to look at in regards to your brew water. If you refer back to the comparison of Madison to Seattle water you'll see that a lot of the compounds listed are dissolved substances in the water. There are a couple though that are more a reference to the molecular side effects of the various things that have been dissolved into or taken out of the water due to the water source and/or the filtration system: Chlorides and pH Level.

The presence of chlorides in water will (over the long term) cause metal to corrode. It is a little different than aggressive or low TDS water breaking metal down due to its solvent nature. Corrosion is a chemical reaction that starts to pit the metal. Eventually you will get a pin-hole leak at a weak point, usually at a weld, and then a leak. RO systems are effective at removing chlorides as are many ion-exchange processes.

pH is the measurement of how acidic or alkaline a liquid is. Water (pure water) is considered neutral with a pH level of 7. Specifically, pH is a reference to the concentration of hydrogen in the liquid. The lower the pH the more acidic it is. In water, the pH level changes based on the various chemical processes (both natural and employed by filtration systems) it is exposed to. pH level is one predictor for how the water will react when in contact with other substances. Since we run water through all sorts of metallic and plastic materials inside of coffee brewing equipment, not to mention that we soak ground coffee in it, pH is an important number to consider. Generally, water should be within 1pH of 7 or neutral. Since some of the ion exchange processes load hydrogen into the water it is important to know the pH of your water.

Final Thoughts
At this point you're probably thinking you'll never boil another gram of water again. You could always cold-brew everything and just forget about it (did you also forget that water has to be hot to brew good coffee? can't do it with cold water), but where's the fun in that? Hopefully, you know a little more about what water does when heated and what will happen if certain levels of dissolved solids are too high or too low. The first thing you should do is taste your water. Don't drink it, taste it. Seriously. Go pick up a gallon of distilled water and do a tasting against your tap water, both at room temperature. Distilled water is at 0 TDS and you will taste the difference right away. The second thing you should do is test your water. Test the water before it goes through any kind of filter and test the water where it hits your brewing equipment. Most Universities will even do it for free or minimal cost as part of the lab services. Armed with the knowledge of what is and isn't in your water, you can start to figure out what it is you should do. RO is definitely a great way of cleaning and grooming your water for good coffee brewing but it tends to be expensive, require regular maintenance, and you need to be careful of the calcite cartridge and what specifically is going back into the water. Ion-exchange is also effective and I like the idea of not having to waste water to get water. But, again make sure you know the effect the ion-exchange is going to have on the other attributes of the water. Also, CHANGE YOUR FILTERS. That sounds obvious, but whether you are going with RO, Ion-Exchange cartridges or tanks, coarse filtration and carbon you need to change the filters, cartridges,membranes or substances etc. involved in the process. Test your water on a regular basis so you know what is happening and when. In the end the list below is a good 'recipe' to go with; in my opinion it is a water that will boil and heat nicely and not break your equipment down while brewing fantastic coffee.

TDS = 50-200ppm
Hardness = 3-5grains
Chlorides = 0
Chlorine/Iron or other taste/odor causing agents = 0
pH = 6-8

If I had my way I'd set up a few filtration systems up on both batch brewing and espresso machines to taste specific TDS and mineral levels and the difference each has on the respective brewed cup of coffee or espresso. I also think it would be fun to cup one coffee with varying amounts of TDS from 0 up to RO waste water (never tested it but very hi ppm I'm sure). It would be such a great lesson in TDS's effect on coffee flavor. Use the same coffee, brew form and brew recipe and just vary the water used. Any way you look at it, water chemistry is probably the most important variable in the extraction of coffee. I'll end by stating a few of my untested assumptions and anecdotal thoughts that, whether true or not, demand that we give just as much thought to our water system as we do our coffee roaster, espresso machine, and brewing system.

The lower the level of TDS the more we can taste inconsistencies and problems with grind, dose and water-temp.

Brewing with low levels of TDS will cause the acidity of the coffee to overpower other attributes.

Generally, brew times can be faster with lower levels of TDS in water.

With water at 50-200ppm levels of TDS it is easier to brew a balanced (body/acidity/flavor) cup.

Softened water and high (+200) levels of TDS in water brew a masked version of what the coffee could be. Its like looking at the world through fogged glass.


  1. Your post implies you need 17 bar inlet pressure to produce RO water. It depends what osmotic pressure is to be overcome and that is a function of the TDS of the feed and required product water. Common small scale (10") membranes can produce 100 gallon per day from only approx 3 bar. And there are membranes that produce more and less with more and less pressure.

  2. Yah I guess it does... Certainly, anything greater than osmotic pressure will work but I haven't seen a system that uses line pressure (3 bar-ish) be very efficient. The combination of membrane design, pre-treatment, specific incoming water chemistry will all effect it. You have any recommendations for a good RO system?

  3. This is a great blog. I will definitely visit again. You should take a look at this site - I use them all the time and have never experienced any issues. Waste Water Testing

  4. Thanks for some other fantastic post. The place else may just anybody get that type of info in such an ideal means of writing? I have a presentation subsequent week, and I am at the look for such information.AGS Filtration Pvt. Ltd is a Native Indian organization of Worldwide reputation in the field of multiple demand and program Filtration. Liquid Filters.

  5. It is truly a great and helpful piece of information.
    I am satisfied that you simply shared this useful information with us.
    Please stay us informed like this. Thanks for sharing. I know something information, to know you can click here

    water cooler
    water delivery