Sidney A. Taylor, Daniel P. Werner
CRC-Evans Rehabilitation
Systems Inc.
Houston
Removing old or disbonded pipeline coatings, preparing surfaces for recoating, and applying multicomponent liquid coatings can be accomplished over longer stretches of pipe and at reduced cost with over-the-ditch technology that uses water and air or mechanical abrasion.
In jobs for United Texas Transmission Co. and Natural Gas Pipeline Co. of America, a proprietary system of high pressure water-jet blasting for cleaning and air abrasive or mechanical-wheel blasting for surface preparation increased productivity and decreased time required for long-line pipeline rehabilitation.
This first of two articles on the process for cleaning, conditioning, and coating pipeline by line-travel equipment details the equipment and procedures and notes tests of the process for asbestos-containing coatings.
The conclusion (OGJ, Feb. 14) presents the field results of the process' use on the projects for both United Texas Transmission and Natural Gas Pipeline.
COATINGS' RESPONSIBILITY
Refurbishment of existing pipelines has become increasingly important for many operating companies because original coatings, especially those 20-30 years old, no longer adequately protect against corrosion.1
Corrosion ranks second only to human error as a cause of pipeline failure.2 Because coatings are the first line of defense in protecting pipelines against corrosion, they must be well bonded, continuous, and resist effects of their environments.
If existing coatings fail in this responsibility, owners are faced with the prospect of either replacing the pipeline or rehabilitating the pipeline's coating.
If a pipeline coating should need to be rehabilitated, several factors, including the following, must be evaluated:
- The type and condition of the original coating
- Whether the pipeline can be taken out of service
- The conditions under which the pipeline must operate, such as service temperature and existing soil conditions.
Depending on these factors, owners have several options, including use of liquid-applied coatings.
In the past, liquid-applied, plural-component coatings were considered only for rehabilitation of short sections of pipe which some owners accomplished while the lines were in service.
Now, however, liquid coatings can be efficiently applied to long sections of large-diameter pipe as well. Such coatings for pipeline rehabilitation offer several advantages, including good resistance to soil stress and cathodic disbandment.
An important consideration in any rehabilitation project is the anticipated cost, and a major factor in cost analysis is the time required to do the work.
Discussed here are the anticipated performance for removal of existing coatings by use of high-pressure water jets, surface preparation by air-abrasive blast and mechanical-wheel blast techniques, and application of plural-component coatings.
ANTICIPATED PERFORMANCE
Among several variables coating-removal performance depends primarily on the coating in place. The three most common coatings encountered in the U.S. are coal-tar enamel (CTE), asphalt, and tape.
Although the behaviors of CTE and asphalt are fairly predictable, the several factors that affect the rate of cleaning them from the line include the following:
- Size of the fine
- Hydraulic horsepower of the high-pressure pump unit
- Thickness of the coating
- Amount of coating disbondment
- Degree of adhesion of the remaining coating
Surface preparation carried out before the original coating - Presence of fiber glass reinforcing in the outer wrap 0 Ambient temperature.
More than 600 miles (1,000 km) of pipeline have been cleaned with proprietary high-pressure water jets (Fig. 1).3 The effects of line size and of hydraulic horsepower can be eliminated by considering cleaning rates in terms of square feet/minute/hydraulic horsepower.
Previous work has shown that cleaning rates should be in the range of 0.15 to 0.29 sq ft/min/hydraulic hp (0.014-0.027 sq m/min/hydraulic hp) for CTE and asphalt systems.45
Coating-yard cleaning tests conducted on a well-adhered CTE coating resulted in a cleaning rate of 0.15 sq ft/min/hydraulic hp (0.014 sq m/min/hydraulic hp), a rate within the expected range.6
Based on these data, cleaning rates for in situ rehabilitation can be predicted in the range of 20-40 sq ft/min (1.9-3.7 sq m/min), and 75-150 sq ft/min (7-13.5 sq m/min) for a large line-travel cleaning system.
Removal of tape is more difficult to predict.
In addition to line size and hydraulic horsepower of the high-pressure pump unit, removal is also affected by the number of tape layers, amount of tape disbondment, amount of active butyl present, degree of surface preparation prior to application of the tape system, and presence of primer under the tape system.
Previous work with removing tape systems by high-pressure water jets has shown a wide variation in cleaning rates. On average, production rates of 50% of those for CTE can be expected.
Rates as low as 25% have been encountered when a great deal of active butyl was present; rates as high as 100%, when the tape was largely disbanded.
Consequently, cleaning rates for the in situ rehabilitation system should be between 5 and 20 sq ft/min (0.5-1.9 sq m/min), and 19-150 sq ft/min (1.75-13.5 sq m/min) for a large line-travel system.
Field use of the equipment, as described in the concluding article of this series, has proven this prediction.
REMOVING ASBESTOS
Until recently, most asbestos-containing materials were removed manually (hand chipping or scraping and hand-held water lances). In addition to being very costly (up to $400/ft, or $1,325/m), manual removal methods are quite slow and cumbersome.
A typical manual-removal process, utilizing three certified technicians, would achieve a cleaning rate of approximately 4 ft/day for each worker on 30-in. pipe (1.22 m/day/worker on 762 mm).
While this may be acceptable for bellholes or short segments, this approach is uneconomical for removal of asbestos-containing materials on long segments.
With the assistance of the Gas Research Institute (GRI), Chicago, equipment and procedures have been developed for removal, collection, transfer, and containerization for safe disposal of asbestos-containing material from line pipe.
The scope of this R&D project included review of applicable regulatory requirements, design and fabrication of equipment, and yard and field experiments.
With high-pressure water jetting equipment, production rates of 45 sq ft/min on pipe sizes 6-16 in. (4.2 sq m on 152-406 mm) and 60 sq ft on pipe sizes 18-36 in. (5.6 sq m on 457-914 mm) are anticipated.
These production rates are realistic and comply fully with U,S. Environmental Protection Agency (EPA), National Emissions Standards for Hazardous Air Pollutants (Neshap), and Occupational Safety & Health Administration (OSHA) regulations for wet-method removal and no visible emissions.
Air monitoring is a continuous process to ensure compliance with the EPA action level of 0.1 fibers/cu cm. Additionally, staged filtration of the effluent water through a 5-mu filter allows for discharge to earth (with approval from appropriate agencies).
It is expected that a large percentage of the water used in this process will be filtered sufficiently to allow for discharge onto the right-of-way.
CALIFORNIA TEST
A field experiment utilizing the asbestos-containing material removal system was conducted in the California desert in 1992.
The scope of the experiment was to remove an asphaltic coating with an asbestos-containing outerwrap from 6,000 ft of 16-in. pipe (1,829 m of 406 mm).
This work was performed in situ with the line operating at reduced pressure. Before the work was initiated, a thorough safety analysis, including finite element analysis of the pipe under load, was conducted.
Even though the coating was an asphaltic-based material, analysis of the effluent water showed it to be of high quality and purity. No traces of petroleum hydrocarbons were found in the effluent.
Sampling for volatile organic compounds (VOCS) revealed nondetectable levels of all VOCs except toluene and xylene, whose highest readings were 7.0 and 3.5 mu liters, respectively, and well within EPA limits.
Approval for discharge of the effluent to earth was therefore sought from and granted by the California Regional Water Quality Control Board and the respective county health departments.
The effluent water for which discharge permission was granted had been recycled through the cleaning system several times. Approximately 30 gpm were recycled through the system's over 4,000 ft (114 l./min over 1,219 m) of cleaning operation.
This water was recycled from the filters through the cleaning system until chloride levels reached 200 ppm, at which point it was diverted to a holding tank, pending analysis, to avoid the risk of chloride contamination on the pipe. Freshwater was then introduced into the system.
Air-monitoring results were all less than the EPA's action level 0.1 fiber/cu cm. The cleaning quality and efficiency of the water-jet process used on nonasbestos-containing material projects were maintained.
All corrosion cells present on the pipe were cleaned to the point that inspections and repairs were easily made. The equipment was fully able to clean the pipe, containerize the asbestos materials, and purify the effluent water in one pass, at temperatures ranging from 30 to 110 F. (0-43 C.).
During the course of the experiment, removal rates averaged 10 fpm, or 42 sq ft/min (3 m/min, or 3.9 sq m/min).
The encouraging results of this field experiment led to the continuation of the research and development project, and the asbestos-removal system is now complete. This system provides an option that significantly reduces the cost and time required to conduct long-segment, asbestos-containing material pipeline rehabilitation.
SURFACE PREPARATION
Correct surface preparation is generally recognized as the single most important factor in success or failure of a protective coating application.78
In theorY, barrier-type coatings protect the substrate by providing a high electrical-resistance path that impedes ion movement in any potential corrosion cell.
All paint coatings, without exception, allow the passage of some water and oxygen through the film. Amounts vary according to thickness, natural permeability, and integrity of the applied film.
The ability of the coating to slow the movement of ions is considerably improved if it resists permeation and creep at the coating substrate interface.
AIR-ABRASIVE BLAST
Although not the most efficient method for every application, air-abrasive blast cleaning is a proven and well-established method of surface preparation. Fig. 2 shows an example of an air-abrasive blast cleaning machine.
Because the pipeline has already been cleaned by high-pressure water jets, surface preparation performance is fairly predictable.7 Nevertheless, several factors which still exert strong influence include the following:
- Pressure and air flow of the air compressor system
- Condition (grade) of the pipe to be blasted
- Amount of primer stain left on the pipeline
- Type of abrasive used
- Abrasive particle size and consistency
- Flow rate through the nozzle (rich vs. lean)
- Pressure and nozzle size
- Nozzle stand-off distance and angle of impingement
- Type of nozzle
- Anchor profile required.8
Few objective cleaning-rate data have been reported for the numerous abrasives commonly used today. The exception is the cleaning-rate data for silica sand.9
The relationship between nozzle diameter, surface preparation grade, and steel condition is shown in Table 1. The cleaning rates shown are generally, considered to be the maximum rates achievable under optimum operating conditions.
Some limited data have been published on the comparative cleaning rates of Staurolite, coal slag, copper slag, and silica sand at various air pressures.10
Although a few specific data points seem inconsistent and the authors have never firmly established the surface-preparation grade obtained, some general conclusions can be drawn about the relative cleaning rates compared to silica sand: staurolite had an improvement in cleaning rate of 113%; coal slag, an improvement rate of 16%; and copper slag, of 30%.
A model predicting actual production rates based on pipeline size, initial pipe condition, final surface finish required, nozzle size, and nozzle quantity has been reported.6 With this model and the configuration of the in situ rehabilitation system, the anticipated performance of this equipment is shown in Table 2.
MECHANICAL-WHEEL BLAST
The amount of work done by a mechanical-wheel blast unit (Fig. 3) is calculated from the equation for kinetic energy 11 12: E = mv2/2; where: m = mass and v = velocity of the abrasive particles.
An increase in velocity effects a very significant increase in the kinetic energy of the particle. Velocity is governed by the combination of wheel peripheral speed, wheel diameter (inside and outside), and shape and length of the wheel blade.
Kinetic energy is also directly proportional to the mass of the particle. Consequently, the amount of work that can be performed is directly related to the amount of metallic abrasive that can be hurled at the surface per unit of time.
All major manufacturers of centrifugal blast wheels rate their wheels' performances in terms of square feet of cleaned surface per horsepower per minute. Manufacturers consider this relationship to be linear.13
Based on the authors' experience, the performance that can be expected for mechanical-blast wheels' cleaning a curved surface, such as a pipeline, is shown in Table 3.
From these data, the anticipated performance of a specific mechanical-wheel blast device can be determined.
A sample calculation is provided in Table 3 for cleaning a 26-in. (660-mm) pipeline on Grade C material to a NACE No. 2 (near-white blast-cleaned) surface preparation grade with a 200-hp unit.
Coating application, with automated line-travel equipment (Fig. 4), is limited by the travel rate of the surface-preparation equipment. Consequently, the anticipated performance is the same as that for surface preparation.
REFERENCES
- Pfaff, T.J., "Requirements for External Pipeline Refurbishing Coatings," presented to the 1989 Pipeline Rehabilitation Seminar, Houston.
- Report No. NTSB-PSS-78-1, U.S. National Transportation Safety Board, Washington, D.C. 3, Taylor, Sidney A., and Chapman, Gerald Cleaning Pipelines Using High Pressure Water jets," presented to Corrosion/91, Houston.
- Frenzel, L.M., "Application of High Pressure Water Jetting in Corrosion Control," presented to the 1995 SSPC Annual Symposium, Pittsburgh. 5, Frenzel, L.M., "Evaluation of 20,000 psi Water Jetting for Surface Preparation of Steel Prior to Coating," presented to the 1995 SSPC Annual Symposium, Pittsburgh.
- Taylor, Sidney A., "An Engineered Approach to Cleaning, Surface Preparation and Coating for Short Segment In-Situ Rehabilitation Work," presented to the 1991 Second European and Middle Eastern Pipeline Rehabilitation Seminar, Brussels.
- Taylor, Sidney A., "Surface Preparation and Application of Plural Component Coatings to Pipeline During Rehabilitation," presented to Corrosion/91, Houston.
- Bayliss, D.A., "Surface Preparation - The State of the Art," presented to the SSPC Annual Symposium, Pittsburgh, 1985.
- Pauli and Griffin, Abrasive Blasting Performance Guide, Vacaville, Calif.
- Seavey, M. "Abrasive Blasting above 100 psi," journal of Protective Coatings and Linings, July 1985, pp. 26-37.
- Mallory, A.W., "Mechanical Surface Preparation: Centrifugal Blast Cleaning," Good Painting Practice, Steel Structures Painting Manual, Vol. I, 2nd ed. (Pittsburgh: SSPC, 1982), pp. 22-31.
- Borch, E.A., "Metallic Abrasives," Good Painting Practice, Steel Structures Painting Manual, Vol. I, 2nd ed. (Pittsburgh: SSPC, 1982), p. 35.
- Technical Bulletin, Vol. VIII, No. 8, December 1988, Ervin Industries, Ann Arbor, Mich.
Copyright 1994 Oil & Gas Journal. All Rights Reserved.
