In response to requests from Plumbing Manufacturers Intl. (PMI) and its members, as well as from other supporters of the U.S....
Residential reverse osmosis (RO) drinking water systems can significantly improve the quality and flavor of most potable drinking water supplies. Residential RO system styles include, among many others, countertop models, low-output faucet-attaching models and the traditional under-the-sink systems found in an ever-increasing number of homes.
With the bulk of the mineral content removed, water performs nearly every rudimentary task more effectively. A few of the in-home uses for RO-treated water include cooking water, auto windshield washer fluid, ice cubes, mixed beverages, and my favorite, incomparable drinking water. When properly sized, applied and maintained, RO will provide unparalleled results and customer satisfaction. To ensure this, let’s consider what can go wrong and how to avoid those unexpected traps that lead to performance failures.
RO can substantially improve water quality in so many ways, and for so many purposes, that it is completely unnecessary to exaggerate the benefits. RO systems, however, must not be sold as having the capability to make an unsafe influent potable. These treatment systems may be able to reduce levels of potentially harmful contaminants, but they are generally not designed to reject these constituents to the levels required to make an otherwise questionable or unsafe water supply safe for potable use.
Drinking plenty of purified water is certainly beneficial to good health, so avoid making unsubstantiated claims that may tend to mislead consumers. Maybe the most important benefit of RO is how the product itself naturally promotes drinking greater amounts of water. This results in increased consumption of healthy water and decreased consumption of other less-beneficial fluids.
Knowing and accounting for the feedwater characteristics is critical when considering an RO system. Pressure, temperature, sediment, hardness, chlorine/chloramines, iron/manganese and total dissolved solids (TDS) are some of the most important parameters to consider. Ignoring or taking for granted any of these factors can lead to reduced production quantity, insufficient rejection performance and a compromised longevity of the system. Take care of the essentials, and your reward will be a quality product that makes plenty of great-tasting water.
RO systems depend on pressure that is in excess of the combined back pressure from the TDS plus the storage tank in order to produce water. A minimum net pressure of about 30 psi is required to attain satisfactory results. To calculate the net pressure available, start by accurately measuring the incoming pressure, preferably at the inlet of the membrane. From this figure, you will subtract 1 psi for every 100 ppm of TDS to account for the osmotic back pressure. Additionally, subtract the initial storage tank pressure.
For example, an incoming pressure of 45 psi, with 800 ppm of TDS and an empty storage tank pressure of 6 psi, would calculate like this: Start with the 45-psi influent, subtract 8 psi for the TDS, and subtract an additional 6 psi for the storage tank. This gives you 45-8-6 = 31 as the net available pressure. Technically, the 31-psi result may be adequate. Remember, however, that the 30-psi barrier is not a magic number where anything above is fine and any result below will not work. Using a booster pump for low or marginal net pressure installations to increase the incoming pressure will result in improved rejection and increased production.
RO membranes are typically tested and rated based on a laboratory temperature of 77˚F. This may be convenient for the testing facility, but it is not the water temperature generally encountered in field installations. Consider that a membrane producing 50 gal per day (gpd) at 77˚F will produce only about 30 gpd at 50˚F, even when all other factors remain constant. These oftentimes unrealistic production estimates can be extrapolated into complete system claims. These figures are not necessarily incorrect or even misleading, but they need to be kept in perspective when explaining expected performance to the end user. Based on reduced temperature alone, it is not unusual to find an anticipated open flow of 2 gal per hour (gph) actually flowing at only 1 gph.
Typically, a nominal 5-micron sediment filter is adequate to reduce suspended matter prior to the membrane. This is commonly followed by a chlorine-removing carbon block filter, as thin film composite (TFC) membranes are very sensitive to chlorine exposure. If using a cellulose triacetate membrane, the influent must be disinfected, so a carbon prefilter would not be used.
Next is the heart of the matter, our membrane. All other components of an RO system are designed to protect or supplement the membrane. Following the membrane can be an optional slow-flow extended-contact inline granular activated carbon filter. The relatively low permeate flow from the membrane makes this filter especially effective at reducing low molecular weight organics. Water is then directed to the storage tank, where it awaits demand from a faucet or other source. Upon exiting the storage tank, water is normally directed through an inline or cartridge granular activated carbon filter to pick up any lingering taste or odor. If desired, the water can subsequently be treated by an ultraviolet system. If complete TDS removal is desired, a mixed bed deionization filter can be used to reduce the TDS to nearly zero. Highly dependent on the TDS level and quantity of the permeate, the longevity of these filters is often inadequate for practical use. Calcium carbonate post filters are becoming more popular. These will add mineral hardness back into the final product, producing what many consider to be a pleasant flavor.
Watch out for membrane killers—contaminants in levels that will foul, hinder or interfere with good RO water production. Chlorine and chloramines are very detrimental to TFC membranes and must be substantially removed to prevent premature failures. Hardness should not exceed about 7 grains per gal, though the lower the hardness, the better. RO membranes more easily reject the added sodium from softeners than the calcium hardness in hard water, so the lower the hardness, the longer you can expect the membrane to last.
Iron at or above about 0.3 mg/L and manganese at or above about 0.05 mg/L will likewise be problematic to the membrane and overall performance of the system. TDS is a variable parameter that incrementally increases osmotic back pressure on the membrane as the TDS level rises. Many manufacturers use an arbitrary number like 1,000 mg/L as an operating limit, but this is really dependant on several factors. The 1,000-mg/L TDS level works as a good guideline for the majority of installations, so it is a natural starting point. Keep in mind the previously mentioned minimum net pressure requirements and how to determine TDS limitations and pressure requirements.
RO is used for standard under-the-counter residential applications requiring just a gallon or two per day, all the way up to industrial desalination plants for entire communities producing multiple tens of thousands of gallons per day. This cost-effective method of processing water can turn a desert into an oasis, or stinky, unappealing tap water into a highly desirable beverage. My son literally cringes when drinking water from the faucets at school. He told me that after football practice, he ran out of water brought from home, but in desperation, he was able to drink a small amount of school water and temporarily quench his thirst.
Like so many of us, he is accustomed to the fresh, clean taste of water processed by the RO method. We all have to hydrate, so it might as well be an inexpensive and pleasurable experience.