[modula id=”4206″]

Improper Filter Installation

Installing a water filter for any larger piping system is often a capitol decision costing $50,000 or more.  In most cases, it is in direct response to a recognized corrosion problem – where its function may play an even greater role in protecting millions of dollars of equipment, product, revenue, and/or infrastructure.

For any heavily corroded system, substantial rust deposits will line the pipe interior to both eliminate any further benefit from the chemical treatment program, and to potentially interfere with system operation, mechanical equipment, and heat efficiency, etc.  From our experience, removing heavy rust deposits is an absolute necessity to reducing corrosion; with it impossible to slow corrosion otherwise.  Contrary to the claims of most chemical treatment suppliers, no protection is provided to the base steel beneath 1/2 in. or more of iron oxide tuberculation.  And as a result, corrosion continues unabated.

While removing such internal deposits is the fundamental objective of all filtration systems, it is very rarely achieved due to a combination of some very common errors.

• A Common Misinterpretation

In reality, what often occurs is that the filter removes a certain minor volume of suspended rust, dirt, and particulates within the first few weeks of operation, but then slows dramatically in its backwashing cycle or cleaning frequency.  This change is then often misinterpreted that the corrosion problem has been resolved.  The filter selection and installation is viewed as a total success, and the problem is forgotten.

In other cases, the reduction in particulate capture is misinterpreted as a failure of the filter unit’s operation and criticism raised with all involved.  Both interpretation of events, however, are wrong.

In most examples, a filtration unit, and especially a high efficiency sand filter or basket/bag housing type filter, will clean up the fine suspended particulates and turbidity of the water to produce a noticeable improvement in water clarity, color, and appearance.  Before and after water samples are often dramatic proof that the filter works.  This improved water quality, in turn, is incorrectly perceived as an improvement in removing deposits from the piping also – which is very rarely the case.

For a piping system of any reasonable size, years of high corrosion activity can produce substantial internal deposits.  A corrosion rate of 5 MPY, very common today for any open condenser water system, will remove approximately 64 lbs. of pipe steel per year per 100 linear feet of 12 in. schedule 40 pipe.  Oxidizing to approximately 12 times its original volume, this loss of pipe steel is transformed into approximately 1.6 cu. ft. of iron oxide – per year, per 100 ft. of pipe.  Extrapolated over the full size of a typical condenser water system and its quite easy to estimate literally thousands of pounds of rust product still attached to the pipe walls.

A quick look at our Corrosion Photo Gallery illustrates this extremely common pipe condition.  Some pipe sections illustrated, in fact, had high efficiency filtration units operating.

• Basic Laws Of Physics

The first problem associated with virtually all water filter installations and a primary reason for their ineffectiveness and low particulate removal is related to physics.

According to Newton’s first law of physics, a mass in motion will continue in its same direction unless acted upon by a force.  The law of inertia is another definition of the same principle.  For water traveling through a piping system, various rust particulates may exist having a wide distribution in physical size and weight – from a microgram to one pound or more.  A preliminary investigation to most filter selection is often a particle size distribution study, although they are all fundamentally in error by only addressing those particulates suspended in solution and in motion at the time of sampling.

Particle size distribution studies, which are always performed as a selling point to high performance filtration units, typically define the bulk of a rust problem in the under 25 micron range.  This result, however, ignores the larger rust deposits on the pipe surface which could be captured with a tennis racket – as illustrated at left from a client’s condenser water pump strainer.

Rust attached to the pipe interior, deposits along the bottom of the lines, and sediment in tower basins and captured in strainers, the overwhelming majority of the actual corrosion problem, are not counted – thereby greatly altering the understanding and perception of pipe condition in terms of rust volume.

Depending upon flow velocity, any suspended and moving rust particles will continue in the same direction.  Unfortunately, most take-off locations for the water filter inlet is perpendicular to flow, thereby requiring the rust particles to stop in motion, turn 90º, and move perpendicular into the filter inlet – a virtual impossibility for any larger particle.  Only those particles of smallest size and mass are capable of being influenced by the suction flow at the side inlet for capture.  Larger particles having too great a mass to alter their direction continue forward past the filter.

Most often, side stream filtering units are installed across the suction and discharge headers of the pumps in order to utilize their 40-80 psi differential to move the water through the filter without an additional pump, and again with the take-off point perpendicular to flow.  Since almost all pumps have coarse strainers at their inlets, largest particulates may be captured there, or shattered into smaller particulates for re-suspension into the system.

We illustrate this very common installation problem in the photographs below.  In the top left photo, the inlet take-off to the side stream filter is actually at the inside radius of a 90° elbow, with the water flow direction approaching from the rear left.  Any rust particulates would be forced to the opposite side of the pipe at right and glide along its outward radius wall – never even reaching in proximity of the filter inlet.  No possible benefit of this water filter is therefore realized, as it exists in name only.  An identical condition is shown at top right at another refrigeration plant.

Middle left and right, we show another common filter take-off at the 3 o’clock position, better of course than the previous two examples, but unlikely to capture rust particles flowing through this 12 in. chill water line.  At longer horizontal lengths, water becomes laminar rather than turbulent to further remove the movement possible bringing a rust particle to the take-off filter suction.  With any larger rust particle or deposit falling by gravity to move along the bottom surface of this pipe, such side capture ports are almost totally ineffective.  And yet, it is the most common installation design we see.

Bottom left shows a take-off from a chill water system actually installed at the inside radius of an elbow.  The arrow shows the direction of flow.  Any particulates within the flow would be forced by inertia to the outside radius similar to the principle behind centrifugal separators, and entirely avoid discharge to the filter.  A worse placement within the entire chill water piping system would not be possible.  At bottom right, a top exit port following another elbow again illustrates the common errors in installation.

The very common scenario is that the water filter, promoted and sold based upon claims of sub-micron retention, will actually clean the moving stream of water of its smallest size and smallest mass particulates capable of being captured by the filter inlet and commonly responsible for the visual turbidity and cloudiness of a fouled piping system.  Larger particles continue on past the inlet point to settle elsewhere and never reach the filter, while all attached deposits remain stationary at the pipe interior to continue their aggressive under deposit pitting.

Turbidity of the water caused by its smallest moveable particulates is removed due to the sub-micron filtration unit, suggesting a successful resolve of the problem by the new filter, while in reality, 98% of the internal rust deposit problem, by weight, still exists.  As we have documented in many investigations where expensive filtrations units have been recommended in response to a major corrosion condition, they have only removed the smallest diameter suspended particulates which pose no threat to the facility, while leaving the overwhelming majority of the problem solidly attached to the pipe wall to continue high under deposit corrosion.

• Proof Of Low Particulate Removal Efficiency

From our extensive involvement in ultrasonic pipe testing, we have well documented the volume of iron oxide and other particulate deposits which commonly accumulates in condenser water and other HVAC piping systems.  Given even a low corrosion rate of 1-2 mils per year (MPY), hundreds and possibly thousands of pounds of rust deposits can accumulate over one or two decades, and will remain attached to the pipe without some mechanism to remove them.  A not uncommon 20% loss of wall thickness at 12 in. Schedule 40 condenser water pipe actually means that 10.7 lbs. of steel has been lost per every linear foot, every year.

For an 800 ft. riser system, that same 20% thickness loss means that approximately 8,560 lbs. of steel has been oxidized and transformed into a less dense mass of iron oxide having approximately 12 times its original volume.  Typically, some percentage of that volume will still exist within the piping system to cause further and more aggressive under deposit corrosion.

Any expectation of removing such high internal corrosion product, or even a small proportion of it through the use of a side take-off intake to any by-pass filter, is virtually impossible.  With almost all filtration units installed where convenient, rather than where most effective, at most, such an installation will only provide the most minimal removal of the smallest suspended particulates, leaving the overwhelming balance behind.  An improperly installed filtration unit will not even keep up with the rate of new rust deposits naturally created.

Were the removal of such a high volume of debris, scale, rust, and other particulates effective, new problems would be created at the drain lines where the backwash was directed.  Our experience has shown only some automatic filter installation to include a settling tank to collect the backwashed deposits, with many backwash lines piped directly to drain.  Yet, clogged floor drains and sumps, which would occur if hundreds of pounds of heavy rust deposits were directly discharged into them, rarely, if ever occur.

In one way, piping the backwash directly to drain eliminates any evidence of filter efficiency, but by definition of not creating drain line problems, proves that the filter is not as effective as likely planned.  The below photo, taken of an automatic side stream filter installation having a collection and settling tank for backwash deposits, is a prime example.

Shown at left and viewed from a downward inside perspective, this 3 ft. diameter x 4 ft. high settling tank had collected so few deposits over a 1.5 year period that not even the entire bottom of the settling tank was covered.  The filter, one of the most effective automatic units on the market, removed perhaps only 1 lb. of rust in over 18 months of 24 hour/day operation of an open condenser water system known to be heavy laden with rust particulates.

Given that the supply and return distribution piping removed from the system were found clogged almost 50% with interior deposits, and that thousands of pounds of particulates were known to exist throughout the 36 floor office property, this high dollar investment in filtration was shown to be virtually worthless.  The purchase of one of the best automatic filters on the market, installed incorrectly at a 2 in. perpendicular take-off to a 18 in. vertical supply riser, ultimately accomplished nothing.

• Chemicals The Second Requirement

A second reason for filtration failure is that the majority of rust volume present will never even leave the pipe surface and therefore cannot possibly be captured or filtered out.

Since years of operation may have deposited hundreds and perhaps thousands of pounds of iron oxide corrosion product at the interior piping surface prior to the filter unit being installed, it is impossible to remove those deposits using filtration alone.  Deposits are typically hardened in place and will resist most mild detergent cleaners.  Tri-sodium phosphate, or TSP, the most commonly used chemical cleaner employed, will only provide mild scouring of the lightest rust particles.

Therefore, every filtration system installed in response to a known corrosion problem requires supplemental chemical agents to loosen and re-suspend the attached deposits back into the water flow and maintain it moving in circulation.  Ideally, efforts to re-suspend the particulates into the water flow should be effective, but not exceed the reasonable capture rate of the filter.

A review of all options with the water treatment contractor should help in choosing a safe but effective chemical adjunct capable of loosening and breaking up the rust deposits and maintaining them suspended in solution.  Removing a sample of deposit laden pipe from the actual system for lab analysis is almost mandatory.  A laboratory trial cleaning of the proposed chemical against those same pipe deposits will even better demonstrate the likely degree of success, and expose any potential damage caused to the pipe itself.

Above all, CorrView International, LLC recommends avoiding any strong acid or alkali cleaners except where thorough ultrasonic and nondestructive testing has been performed to identify any weakened areas of pipe, joints, threads, A/C units, etc.  Corrosion coupons of both steel and copper are especially important during any cleaning procedure in order to monitor for an increase in wall loss.

• Make Use Of Gravity And Inertia

Defining the location and take-off point of the water filter is actually more important than the filter itself.  Instead of a side port connection, a much better alternative is to install the filter at the very bottom of the supply line from the cooling tower, or at the bottom of the downfeed line in the case of a closed system.  This adaptation uses gravity to advantage in not only directing the particulates downward, but with increased force and inertia.  With a higher system pressure normally found at the return piping upward, this configuration requires a supplemental circulating pump at additional installation, energy, and maintenance costs.

The piping layout is critical in itself.  Always possible with proper planning, the filter inlet should be in a straight and downward direction from either the cooling tower or closed system downfeed piping.  This takes advantage of the inertial forces which tend to move all particulates in a straight line whenever the water changes direction.

If possible, we recommend replacing a basement or lowest level downfeed elbow with a full size tee instead, extending the pipe in its full size further downward, and then reducing the tee diameter to the size of the filter pipe inlet.

Extending the bottom of the tee piping in its original dimension approximately 3-4 ft. provides a collection chamber or dead leg, and even further improvement to filtering efficiency.  Expanding this same collection piping to a greater diameter, 6 in. to 24 in. for example, immediately drops the water velocity at this end point – thereby producing greater settling and even more filtration benefit at relatively low cost.

The layout of the piping is also a major factor.  Where low flows and/or long horizontal runs exist, particulates, even if they are loosened by the action of a chemical agent, may not migrate back into the main flow stream to be picked up by the filtering device.  The most effective and sophisticated filtering device becomes virtually useless if insufficient flow exists to move the particulates to its collection inlet.

• Purchasing The Filter Only Half The Issue

In our experience, we rarely find filtering systems installed to provide maximum effectiveness, nor a chemical treatment program even addressing the question of existing interior deposits.  Normally, after spending $50,000 or more on the filtration unit, corners are usually cut to reduce installation and operating costs, and only a small degree of benefit is realized as a result.

Strangely enough, filtration experts fail to maximize the effectiveness of their units once sold by recommending a more effective installation site.  Site selection is typically based upon convenience rather than benefit and lowest cost; with incomplete particulate removal used as the basis to recommend additional filtration units.  In some examples, it is simply incompetence, or the attitude that installation placement does not matter.  A common argument that the filtration unit is so advanced and effective that placement does not matter, is widely used.

While the selection of the filtering device itself is a priority concern, CorrView International recommends providing an equal or greater amount of attention to its placement, installation, and chemical treatment program.  A realistic assessment of pipe quality, volume of deposits, and threat level to the system are also important areas to consider – meaning that ultrasonic and metallurgical testing are a must.

The fabrication of a removable spool piece from the actual piping system is highly recommended in order to assist beginning pipe condition, and its improvement or further deterioration.

Combined, the proper combination of planning and engineering can turn an otherwise poor to mediocre performing water filter into a unit that makes substantial improvement to building or plant operations.

© Copyright 2023 – William P. Duncan, CorrView International, LLC