Epoxy Coating of Pipe Systems by Jack McCuaig, P.Eng.
McCuaig & Associates Engineering Ltd.
The following report considers the use of in-situ epoxy coating of domestic water pipes.
Abstract
Epoxy coating of existing pipe systems while in place is a relatively new process available in British Columbia.
The system is well suited to steel pipes, but has some problems with future repairs to copper pipes.
The epoxy coating process seems to provide adequate pipe coating throughout.
The cost of this system is less than a conventional replacement, but should piping changes be necessary, the cost savings are less.
The Water Pipe Corrosion Problem
A domestic water distribution system is designed to carry fresh water to each fixture in the building. "Recently" (in the last 15 years) the use of plastic piping has become more generally accepted. Although plastic pipe (polybutylene, PVC or CPVC) has some of its own problems, it generally is corrosion resistant. Prior to the use of plastic pipe, however, either copper or galvanized steel pipe was used throughout North America. Some lead, cement or black-iron pipe may be found, but it is rare.
Both copper and steel pipe corrodes, although their corrosion mechanism is quite different. In much of North America, the relatively "hard" water (i.e. high pH) available from wells, lakes or rivers, creates very slow corrosion of copper pipes. Estimates of expected life of a copper distribution in these areas in excess of 150 years, is not unreasonable. In areas of "soft" water (i.e. low pH), the copper piping systems have a significantly shorter life span. Soft water is quite common in the Pacific Northwest and Coastal British Columbia, where the reservoirs are supplied from what is essentially glacial run-off or surface run-off.
Our company has been involved in extensive retrofit work of plumbing systems in the Greater Vancouver area. In this region, copper systems typically have a useful life in the range of 20 years. The soft water causes a general thinning of the pipe wall as well as ongoing pitting corrosion, which results in pin-hoe leaks. The problem is made worse by: heated water (the hotter the domestic hot water is maintained, the shorter the life), undersized pipes causing high flow rates (velocity related pitting), and di-electric corrosion (having steel or iron in contact with the copper pipes). Once the corrosion has progressed beyond a certain level, the system becomes so fragile that shutting down the system to effect a repair causes further leaks necessitating additional shutdowns. This is referred to as a cascading failure.
Steel pipe corrosion is quite different. Galvanized pipe is essentially steel pipe with a layer of nickle based compound coating the steel. The nickle acts as a "sacrificed anode." This means that it dissolves to prevent corrosion of the steel. Unfortunately once the nickle has dissolved or worn away, the steel is left unprotected and rusts. As steel rusts, rather than the pipe-wall thinning (as with copper pipe) the rust builds up inside the pipe. As the rust grows, the pipe becomes blocked. This results in more noise and greater flow losses. Eventually, the pipe would become completely blocked, although generally once the rust has progressed to the point that 50% of the cross sectional area of the pipe is blocked (i.e. the pipe is effectively only half the size), the flow characteristics are poor enough that replacement is necessary. An alternative to replacement recently available in North America is to epoxy-coat the pipes in place.
Epoxy Coating
Epoxy coating of metals, such as storage tanks, has existed for decades. Epoxy is a good material for this purpose for the following reasons:
Once epoxy catalyzes, it is very stable (inert for domestic water).
Epoxy has a tenacious bond to metals.
Although epoxy is considered brittle compared to other plastics, it is far less brittle than the alternatives used for tank coating, such as glass, fibreglass, concrete, etc.
Epoxy has a good hardness tensile and shear strength, making it a tough, resilient material.
Epoxy has quite predictable physical properties (viscosity, cure time, etc.).
The process recently brought to North America involves coating the inside of pipes in-place. It is a very simple matter to coat pressure vessels by brush, towel, or spray application. It is not difficult to imagine the problems in applying an even coating to long lengths of small diameter pipe.
Epoxy Pipe Coating In-Situ
In Europe, there are at least three methods of in-situ epoxy pipe coating. Only one of these methods has been introduced into Canada. Our firm’s experience is with this process. The process was developed in Japan, approximately 30 years ago. This system was introduced into British Columbia in 1997 and initially, there was only one contracting company using this technique. At the time of writing this report, there are two companies performing epoxy pipe coating.
The application technique involves four stages. The first stage is to prepare the metal for the epoxy application. This involves drying the pipe and removing all corrosion by what is essentially sandblasting. High velocity air with an abrasive is circulated through the pipe being treated. Since corroded copper does not build-up on the inside of pipes, it is much quicker to "clean" copper pipes. In fact, if the hard abrasive and time used for steel pipes were applied to copper pipes, the copper would be completely worn away.
Once the pipe has been sand blasted to a clean, bright surface inside, it is heated by circulating warm dry air through the pipe. Temperature of the pipe is critical to assure proper flow of epoxy.
The third stage is to apply the epoxy. A temporary tube filled with pre-mixed epoxy is connected to one end of the pipe. High-pressure air blows the epoxy down the pipe. A clear tube is connected to the other end of the pipe to provide visual assurance that the epoxy has run the entire pipe length. Once this is assured, the flexible tubes are disconnected and the epoxy is left to cure for a minimum of 12 hours.
The four stages (preparation, sand blasting, epoxy application, and curing) require access to both ends of a given section of pipe, as well as the ability to apply high velocity compressed air (and sand) to one end and to remove it from the other. This is achieved by operating a high volume compressor/heater outside the building and running flexible air hoses up the outside of the building. Each pipe section is controlled by switching air valves to provide air only to the location being worked on.
Requirements
There are four basic requirements to be met to fulfill the intent of the upgrade.
Total coverage: All pipes being coated must be coated along their entire length. Bare sections are not acceptable.
Adequate thickness: An even coat must be applied over the pipe length. Pooling of epoxy in areas causing restrictions, or too thin sections, would not be acceptable.
Adequate substrate: In any coating process, the most common reason for failure is inadequate substrate. What this means is that the epoxy must be applied to a properly prepared clean, bright metal surface of adequate strength. All sections of plastic pipe and metal pipe, which have corroded to the point they have lost their structural integrity, must be replaced prior to coating.
Timing: The work generally involves an occupied building. It is generally acceptable to forgo the use of some plumbing fixtures for a short period. Extended time periods without access to water supply for domestic or sewage use is not acceptable.
Potential Problems
Unknown pipes inside the walls: Since the entire pipe is not exposed prior to coating (only the ends being accessed are exposed) it is possible that a section of pipe inside the wall was installed as part of previous repair. Of particular concern would be a section of plastic pipe. The epoxy coating system is not designed (nor is it adequate) to coat plastic pipe. Another problem would be a section of dissimilar metal (a section of steel in a copper pipe system or a section of copper pipe in a steel pipe system). In either case, the surface preparation would be incorrect for the dissimilar section.
In-line Valves: It is possible to inject the epoxy through existing valves. This however renders the valve useless, as it will be set with epoxy into a permanently open position. Valves must be removed prior to epoxy application and re-installed after.
Too Thin Pipe (Copper): If sections of pipe have corroded to the extent that they lack structural integrity, they must be replaced prior to coating.
Undersized Pipe: It is possible that during the original construction, undersized pipe was installed. This may have been by design error, installation error or knowingly "cutting corners." The result is usually poor pressure/flow from fixtures and increased noise. Epoxy coating the pipe will not solve those problems and will likely make them worse.
Incorrect Pipe Size/Length Estimates: During the epoxy-coating phase, the correct amount of epoxy is injected by estimating total surface area to be coated. This means the existing diameter and length of pipe must be known. Incorrect estimates of pipe size/length would result in an incorrect amount of epoxy being applied.
Future Repairs and Renovations: Once a copper pipe has been coated with epoxy, it can no longer be soldered without damaging the coating. Any future repairs or plumbing renovations would require the use of mechanical joint techniques. This presents two problems. First there is a possibility the plumber doing the repair may not know he cannot solder the pipe, and second, mechanical joints are a less dependable repair method. These joints are prone to failure and are not as secure as a proper solder joint.
Case Study
In our capacity as consulting engineers, we were asked to act as consultants for a project of this type. The building was a two zone concrete high-rise.
At the time there was only one application contractor in the province and so tendered bids were not possible. Our role was to establish a performance specification for the project, manage all contracts, payment approvals, provide site inspections to assure conformance to specifications, safety, and to carry-out testing to assure adequate quality of installation.
Twelve pipe samples were removed during the course of work and destructive testing and analysis was carried out in our office. Of particular concern was adequate coverage and thickness of the epoxy coating.
Results of destructive testing are provided in the following tables. For simplicity, average results are presented for each sample. Samples were chosen at random during the project. As can be seen from the "descriptions" in table A, a wide variation of samples was obtained with respect to the amount of corrosion to the pipe prior to coating.
|
TABLE A Pipe Samples | ||
|
Sample |
Type |
Description |
|
A |
Hot |
New pipe – no leaks |
|
B |
Hot |
Soft temper tube – old, but no leaks |
|
C |
Hot |
Heavily corroded multiple leaks |
|
D |
Cold |
Old, but no leaks |
|
E |
Hot |
5 leaks |
|
F |
Cold |
Multiple leaks |
|
G |
Cold |
4 leaks |
|
H |
Hot |
12 leaks |
|
I |
Hot |
No leaks |
|
J |
Cold |
No leaks |
|
K |
Cold |
3 leaks |
|
L |
Cold |
2 Leaks |
*All samples type L copper tubing.
NOTE: Leaks are any pit which has progressed through the pipe wall – it may not have significantly leaked during use due to corrosion blocking the pit-hole until sand blasting re-opened the hole. All samples type L manufactured by Wolverine, except samples E and K, which were type L manufactured by Anaconda.
|
TABLE B Results of Destructive Testing | |||||
|
Pipe Sample |
Diameter |
Remaining Copper Wall Thickness (Average) |
Average Minimum Epoxy Wall Thickness |
Approximate "Ridge" Height |
Approximate "Ridge" Spacing |
|
A |
½" |
0.0435" |
0.025 |
0.093" |
1" |
|
B |
½" |
0.0370" |
0.039 |
0.140" |
5 /8" |
|
C |
½" |
0.0380" |
0.027 |
0.090" |
¼" |
|
D |
¾" |
0.0378" |
0.047 |
0.124" |
7 /8" |
|
E |
1" |
0.0452" |
0.028 |
0.130" |
¼" |
|
F |
¾" |
0.0391" |
0.049 |
0.112" |
¾" |
|
G |
1" |
0.0415" |
0.045 |
0.108" |
1" |
|
H |
½" |
0.0335" |
0.040 |
0.115" |
¾" |
|
I |
¾" |
0.0386" |
0.049 |
0.095" |
1" |
|
J |
¾" |
0.0413" |
0.050 |
0.102" |
5 /8" |
|
K |
1" |
0.0395" |
0.042 |
0.111" |
½" |
|
L |
1" |
0.0438" |
0.041 |
0.107" |
5 /8" |
NOTE: Based on our hoop stress calculations, the minimum thickness allowed for epoxy was 0.025".
Comments on Destructive Testing
Destructive testing revealed that all pipe samples were adequately covered with epoxy to an adequate thickness based on our requirements.
One sample of pipe, which was removed from service due to structural failure, was also provided. The sample was cracked perpendicular to the direction of flow as would be typical of a fatigue failure from stresses due to inadequate allowance for thermal movement.
The crack was noticed at the time of epoxy injection. The section was removed and replaced after the injection was completed.
The epoxy had bridged the crack effectively sealing it. This sample was hydrostatically tested to 200 psi without failure. This would indicate the epoxy has a reasonable ability to repair minor cracks and pinholes.
As can be seen in Table B, the application caused ridging along the interior of the pipe (see photo). This ridging will provide increased structural rigidity, however, of greater concern is the fact that it will cause increased flow losses as well as potential cavitation and increased noise during use.
These concerns were brought to the attention of the building owners. Following completion of the project, there were no complaints received from residents with respect to noise or poor water pressure.
Although it appears that this may not be an issue in this building, a building with "undersized" pipes, which have high existing flow rates, will likely have a greater problem.
Problems Encountered During the Project
As mentioned earlier, the work was carried out in an occupied building. Because of this, a significant portion of our specifications dealt with scheduling and protection of building contents and finishes. Although there were some complaints from residents, with the exception of those related to problems listed below, the number of complaints was in keeping with a successful and well-organized conventional pipe replacement project.
The following problems are isolated incidents and should not be expected in future projects. They are given as an illustration that any new industry is apt to have a few "bugs" to work out. Although this is not a new industry elsewhere, it is new in Canada and as such, some fine tuning may still be necessary.
The most significant problem encountered during the project we studied was the failure of a coupling on one of the main compressed air lines. The line was a 2" diameter reinforced flexible high-pressure hose with quick connect couplings locked in place by cotter pin. The couplings are steel and attached to the hose via a bolt clamping arrangement.
During operation, the clamp released and the hose was then unsecured at one end while high velocity air was running through it. The end then whipped around, freely breaking 4 windows before the compressor could be shut down.
Although no one was injured, the potential existed for very serious injuries from the mishap. Following this, daily checks of hose tie-downs and hose clamps were instituted. This was effective in preventing further failures.
Another problem encountered related to the failure of an epoxy-coated piece of pipe causing a leak. This leak appeared to be due to thermal stress fracture of the copper. The epoxy coating was not adequate to resist the leak. A theory as to why relates to the fact that the pipe was in contact with a heating pipe. This not only amplifies the thermal stresses, but also it affects the epoxy application. Pressure and velocity are adjusted to allow for the expected viscosity ("flowability") of the epoxy at 90° F, the correct temperature of the pipe for application. The pipe in question could have been as hot as 180° F during application, causing a much thinner coating. This likely contributed to the failure of the epoxy to prevent the leak.
It is important to note that the leak was repaired by the epoxy applicating contractor as a warrantee item. The repair was done by their own service plumber. He repaired the pipe by soldering a new pipe into place, in violation of their own recommendations. The pipe was subsequently removed and repaired with a mechanical joint at our request. The above highlights the difficulty in assuring that all future repairs and/or changes do not use solder joints.
General Comments - Epoxy Coating vs. Conventional Replacement
As with any new technique or technology, it should be considered in comparison with previous alternatives. The following comments compare an epoxy coating project with a conventional replacement.
EXPECTED LIFE – The use of in-situ epoxy coating is quite new in this area so it is impossible to provide definitive life estimates. With some reasonable assumptions, however, some estimates can be made. Assuming that the epoxy was adequately applied, and that it will then perform in a similar fashion to epoxy coatings applied by conventional methods, it is reasonable to expect the epoxy coating to remain intact for the life of the building (over 100 years). A conventional pipe replacement should last in the range of 40 years (assuming no water treatment).
DISRUPTION OF RESIDENTS – An in-situ epoxy-coating project will involve less in-suite work, and less cutting open and patching of walls than a conventional pipe replacement project. It will however likely be noisier during the work phase. Residents will be without some fixtures overnight, while in a conventional replacement, all fixtures work every night. As stated previously, epoxy coating involves interior ridging and so flow rate from fixtures are reduced and pipes may be noisier.
SYSTEM CHANGES – A proper design for a conventional pipe replacement would consider any changes that could be made to improve the system. Typically, better "zoning" allowing isolation of smaller zones (e.g. each floor) is provided. Other changes may involve improved recirculation, relocation of supplies, upgrade of boiler room, etc.
These changes can be made with a conventional pipe replacement or an epoxy-coating project. The difference is that for a conventional pipe replacement, these types of upgrades are a minimal cost difference. For an epoxy-coating project, changes like these may be a considerable extra cost.
As stated earlier, future repairs to an epoxy system are not as secure due to the fact that soldering should not be used. Because of this, once a system has been epoxy coated, any changes should be discouraged.
COST – The most attractive aspect of in-situ epoxy coating is that it generally costs between 66% and 75% of the cost of a conventional pipe replacement.
Buildings that require significant layout changes, or have existing problems with undersized pipe and/or noise will likely not realize many savings. For most buildings, however, the cost saving is significant.
The Role of a Consultant
Although a design requiring a sizing of each pipe is not appropriate for this type of work, there are still many reasons to use a consultant.
Scope of Work: Although it is relatively easy to include all existing pipe, generally a project of this type will involve at least some upgrade of existing piping or equipment prior to coating. Options and ramifications must be fully understood by the owners to make an informed decision.
Quality Control: Even when dealing with a reputable contractor, it is always important to assure that the work meets a quality standard. Setting the standard and further on-sit inspections is necessary for this.
Contract Management: Any large contract has a significant amount of management necessary. WCB assurance, statutory lien holdbacks, progress draw approvals, etc. all require a good understanding of construction regulations and/or the actual work being done.
Dispute Resolution: A dispute can be a very minor item or a major disagreement. In either case, it is important to have a professional to act as a referee and/or arbitrator to resolve the dispute in the simples and best manor for all parties concerned.
Liaison with Inspectors: A professional engineer is extremely useful when dealing with municipalities about code issues
Recommendations and Considerations
The most difficult question to answer would be "should my building use an epoxy coating system, or a conventional replacement?"
Every building is different and the owners have different needs and expectations of their system.
Another problem is that the system is still relatively new and without a long track record, it is difficult to endorse any new product or system.
In simple terms, however, the following should be considered.
The application technique has been used for many years in Europe and the United States.
Our limited testing seems to show that adequate coverage of pipes is achieved.
Substantial cost savings can be expected.
If the existing system is to be substantially upgraded or modified, the savings are less significant.
Resident disruption is different for each system and a matter of opinion, which is better.
Future repairs or changes to the system are made more difficult with an epoxy coated system.