PERFORMANCE FAILURES

The failure of heat exchanger equipment to perform satisfactorily may be caused by one or more factors, such as:

(1) Excessive fouling. (2) Air or gas binding resulting from improper piping installation or lack of suitable vents

(3) Operating condftions differing from design conditions. (4) Maldistribution of flow in the unit.

(5) Excessive clearances between the baffles and shell and/or tubes, due to corrosion. (6) Improper thermal design

Improper start-up or shut-down sequences, particularly of fixed tubesheet units, may cause leaking of tube-to-tubesheet and/or bolted flanged joints.

START-UP OPERATION: Most exchangers with removable tube bundles may be placed in service by first establishing circulation of the cold medium, followed by the gradual introduction of the hot medium. During start-up all vent valves should be opened and left open until all passages have been purged of air and are completely filled with fluid. For fixed tubesheet exchangers, fluids must be introduced in a manner to minimize differential expansion between the shell and tubes.

SHUT-DOWN OPERATION: For exchangers with removable bundles, the units may be shut down by first gradually stopping the flow of the hot medium and then stopping the flow of the cold medium. If it is necessary to stop the flow of cold medium, the circulation of hot medium through the exchanger should also be stopped. For fixed tubesheet exchangers, the unit must be shut down in a manner to minimize differential expansion between shell and tubes. When shumng down the system, all units should be drained completely when there is the possibility of freezing or corrosion damage. To guard against water hammer, condensate should be drained from steam heaters and similar apparatus during start-up or shut-down. To reduce water retention after drainage, the tube side of water cooled exchangers should be blown out with air.

TEMPERATURE SHOCKS: Exchangers normally must not be / should not be subjected to abrupt temperature fluctuations. Hot fluid

suddenly introduced when the unit is cold, nor cold fluid suddenly introduced when the unit is hot.

LOCATING TUBE LEAKS: The following procedures may be used to locate perforated or split tubes and leaking joints between tubes and tubesheets. In most cases, the entire front face of each tubesheet will be accessible for inspection. The point where water escapes indicates a defective tube or tube-to-tubesheet joint.

(l) removable channel cover: Remove channel cover and apply hydraulic pressure

(2) Units with bonnet type head: For fixed tubesheet units where tubesheets are an integral part of the shell, remove bonnet and apply hydraulic pressure in the shell. For fixed tubesheet units where tubesheets are not an integral part of the shell and for units with removable bundles, remove bonnet, re-bolt tubesheet to shell or install test flange or gland, whichever is applicable, and apply hydraulic pressure in the shell.

(3) Units with Type S or T floating head: Remove channel cover or bonnet, shell cover and floating head cover. Install test ring and bolt in place with gasket and packing. Apply hydraulic pressure in the shell. A typical test ring is shown in Figure E-4.13-2. When a test ring is not available t is possible to locate leaks m the floating head end by removing the shell cover and applying hydraulic pressure in the tubes. Leaking tube joints may then be located by sighting through the tube lanes. Care must be exercised when testing partially assembled exchangers to prevent over extension of expansion joints or overloading of tubes and/or tube-to-tubesheet joints.

GASKET REPLACEMENT: Gaskets and gasket surfaces should be thoroughly cleaned and should be free of scratches and other defects. Gaskets should be properly positioned before attempting to retighten bolts. It is recommended that when a heat exchanger is dismantled for any cause, it be reassembled with new gaskets. This will tend to prevent future leaks and/or damage to the gasket seating stirfaces of the heat exchanger. Composition gaskets become dried out and brittle so that they do not always

provide an effective seal when reused. flow to match their contact surfaces. Metal or metal jacketed gaskets, when compressed initially: In so doing they are work hardened and, if reused, may prowde an imperfect seal or result in deformation and damage to the gasket contact surfaces of the exchanger.

Bolted joints and flanges are designed for use with the particular type of gasket specified. Substitution of a gasket of different construction or improper dimensions may result in leakage and damage to gasket surfaces. Therefore, any gasket substitutions should be of compatible design. Any leakage at a gasketed joint should be rectified and not permitted to persist as it may result in

damage to the gasket surfaces. Metal jacketed type gaskets are widely used. When these are used with a tongue and groove joint

without a nubbin, the gasket should be installed so that the tongue bears on the seamless side of the gasket jacket. When a nubbin is used, the nubbin should bear on the seamless side.

PLUGGING OF TUBES: In U-tube heat exchangers, and other exchangers of special design, it may not be feasible to remove and replace defective tubes. Defective tubes may be plugged using commercially available tapered plugs with ferrules or tapered only plugs which may or may not be seal welded. Excessive tube plugging may result in reduced thermal performance, higher pressure drop, and/or mechanical damage. It is the users responsibility to remove plugs and neutralize the bundle prior to sending it to a shop for repairs.

SCOPE OF STANDARDS

GENERAL: The TEMA Mechanical Standards are applicable to shell and tube heat exchangers which do not exceed any of the following criteria:

(1) inside diameters of 100 inches (2540 mm)

(2) product of nominal diameter, inches (mm) and design pressure, psi (kPa) of 100,000 (175x106)

(3) a design pressure of 3,OW psi (20664 kPa) The intent of these parameters is to limit the maximum shell wall thickness to approximately 3 inches (76 mm), and the maximum stud diameter to approximately4 inches (102 mm). Criteria contained in these Standards may be applied to units which exceed the above parameters.

DEFINITION OF TEMA CLASS R EXCHANGERS: The TEMA Mechanical Standards for Class R heat exchangers specify design and

fabrication of unfired shell and tube heat exchangers for the generally severe requirements of petroleum and related processing applications.

DEFINITION OF TEMA CLASS c EXCHANGERS: The TEMA Mechanical Standards for Class c heat exchangers specify design and

fabrication of unfired shell and tube heat exchangers for the generally moderate requirements of commercial and general process applications.

DEFINITION OF TEMA CLASS B EXCHANGERS: The TEMA Mechanical Standards for Class B heat exchangers specify design and

fabrication of unfired shell and tube heat exchangers for chemical process service.

CONSTRUCTION CODES: The individual vessels shall comply with the ASME (American Society of Mechanical Engineers) Boiler and Pressure Vessel Code, Section VIII, Division 1, hereinafter referred to as the Code. These Standards supplement and define the Code for heat exchanger applications. The manufacturer shall comply with the construction requirements of state and local codes when the purchaser specifies the plant location. It shall be the responsibility of the purchaser to inform the manufacturer of any applicable local codes. Application of the Code symbol is required, unless otherwise specified by the purchaser.

MATERIALS-DEFINITION OF TERMS: For purposes of these Standards, carbon steel shall be construed as any steel or low alloy

falling within the scope of Part UCS of the Code. Metals not included by the foregoing (except cast iron) shall be considered as alloys unless otherwise specifically named. Materials of construction, including gaskets, should be specified by the purchaser. The manufacturer assumes no responsibility for deterioration of parts for any reason.

U-BEND REQUIREMENTS: When U-bends are formed, it is normal for the tube wall at the outer radius to thin. The minimum tube wall thickness in the bent portion before bending shall be:where

to = Original tube wall thickness, inches (mm)

t, = Minimum tuba wall thickness calculated by Coda rules for a straight tube subjected to the same pressure and metal temperature, inches (mm)

do= Outside tube diameter, inches (mm)

R= Mean radius of bend, inches (mm)

More than one tuba gage, or dual gage tubes, may be used in a tube bundle. When IJ-bends are formed from tube materials which are rafatively non-work-hardening and of suitable temper, tube wall thinning in the bends should not exceed a nominal 17% of

original tube wall thickness. Flattening at the bend shall not exceed 10% of the nominal tube outside diameter. U-bends formed from tube materials having low ductility, or materials which are susceptible to work-hardening, may require special consideration

SQUARE PATTERN: In removable bundle units, when mechanical cleaning of the tubes is specified by the purchaser, tuba lanes should be continuous.

TRIANGULAR PATTERN: Triangular or rotated triangular pattern should not be used when the shell side is to be cleaned mechanically.

TUBE PITCH: Tubes shall be spaced with a minimum center-to-center distance of 1.25 times the outside diameter of the tube. When mechanical cleaning of the tubes is specified by the purchaser, minimum cleaning lanes of l/4 (6.4 mm) shall be provided.

TUBE PITCH: Tubes shall be spaced with a minimum center-to-center distance of 1.25 times the outside diameter of the tube. Where the tube diameters are 5/8 (15.9 mm) or less and tuba-to-tubesheet joints are expanded only, the minimum center-to-center distance may be reduced to 1.20 times the outside diameter.

TUBE PITCH: Tubes shall be spaced with a minimum center-to-center distance of 1.25 times the outside diameter of the tube. When mechanical cleaning of the tubes is specified by the purchaser and the nominal shell,diameter is 12 inches (305 mm) or less, minimum cleaning lanes of 3/16 (4.8 mm) shall be provrded. For shell diameters greater than 12 inches (305 mm), minimum cleaning lanes of t/4 (6.4 mm) shall be provided.

BAFFLES AND SUPPORT PLATES

TYPE OF TRANSVERSE BAFFLES: The segmental or multi-segmental type of baffle or tube support plate is standard. Other type baffles are permissible. Baffle cut is defined as the segment opening height expressed as a the shell inside diameter or as a percentage of the total net free area inside the shell Percentage of shell cross sectional area minus total tube area). The number of tube rows that overlap for multi-segmental baffles should be adjusted to give approximately the same net free area flow through each baffle. Baffles shall be cut near the centerline of a row of tubes, of a pass lane, of a tube lane, or outside the tube pattern. Baffles shall have a workmanlike finish on the outside diameter. Typical baffle cuts are illustrated in Figure RCB-4.1. Baffle cuts may be vertical, horizontal or rotated.

SPACING OF BAFFLES AND SUPPORT PLATES

MINIMUM SPACING: Segmental baffles normally should not be spaced.closer than l/5 of the shell ID or 2 inches (51 mm), whichever is greater. However, special design considerations may dictate a closer spacing.

MAXIMUM SPACING: Tube support plates shall be so spaced that the unsupported tube span does not exceed the value indicated in Table RCB-4.52 for the tube material used.

TUBE BUNDLE VIBRATION: Shell side flow may produce exdation forces which result in destructive tube vibrations. Existing predictive correlations are inadequate to insure that any given design will be free of such damage. The vulnerability of an exchanger to flow induced vibration depends on the flow rate, tube and baffle materials, unsupported tube spans, tube field layout, shell diameter, and inlet/outlet configuration. Section 6 of these Standards contains information which is intended to alert the designer to potential vibration problems. In any case, and consistent with Paragraph G-5, the manufacturer is not responsible or liable for any direct, indirect, or consequential damages resulting from vibration.

IMPINGEMENT BAFFLES AND EROSION PROTECTION: The following paragraphs provide limitations to prevent or minimize erosion of tube bundle components at the entrance and exit areas. These limitations have no correlation to tube vibration and the designer should refer to Section 6 for information regarding this phenomenon.

SHELL SIDE IMPINGEMENT PROTECTION REQUIREMENTS: An impingement plate, or other means to protect the tube bundle against impinging fluids, shall be provided when entrance line values of p I/ exceed the following: non-abrasive, single phase fluids, 1500 (2232); all other liquids, including a liquid at its boiling point, 500 (744). For all other gases and vapors, including all nominally saturated vapors, and for liquid vapor mixtures, impingement protection is required. I/ is the linear velocity of the fluid in feet per second (meters per second) and p is its density in pounds per cubic foot (kilograms per cubic meter). A properly designed diffuser may be used to reduce line velocities at shell entrance. TIE RODS AND SPACERS: Tie rods and spacers, or other equivalent means of tying the baffle system together, shall be provided to retain all transverse baffles and tube support plates securely in position.

SEALING DEVICES: In addition to the baffles, sealing devices should be installed when necessary to prevent excessive, fluid by-passing around or through the tube bundle. Sealing devices may be seal strips, tie rods with spacers, dummy tubes, or combinations of these.

LENGTH OF EXPANSION: Tubes shall be expanded into the tubesheet for a length no less than 2 (50.8 mm) or the tubesheet thickness minus 1 /e (3.2 mm), whichever is smaller. In no case shall the expanded portion extend beyond the shell side face of the tubesheet. When specified by the purchaser, tubes may be expanded for the full thickness of the tubesheet.

LENGTH OF EXPANSION: Tubes shall be expanded into the tubesheet for a length no less than two tube diameters, 2 (50.8 mm),, or the tubesheet thickness minus l/8 (3.2 mm), whichever is smaller. In no case shall the expanded portion extend beyond the shell side face of the tubesheet. When specified by the purchaser, tubes may be expanded for the full thickness of the tubesheet.

TUBESHEET PASS PARTITION GROOVES: Tubesheets shall be provided with approximately 3/16 (4.8 mm) deep grooves for pass partition gaskets.

USE THESE GUIDELINES WHEN WITNESSING A STANDARD HYDROSTATIC TEST ON ANY VESSEL

Vessel configuration

The test should be done after any stress relief.

Vessel components such as flexible pipes, diaphragms and joints that will not stand the pressure test must be removed.

The ambient temperature must be above 0C (preferably 15-20C) and above the brittle fracture transition temperature for the vessel material (check the mechanical test data for this).

The test procedure

Blank off all openings with solid flanges.

Use the correct nuts and bolts, not G-clamps.

Two pressure gauges, preferably on independent tapping points, should be used.

It is essential for safety purposes to bleed all the air out. Check that the bleed nozzle is really at the highest point and that the bleed valve is closed off progressively during pumping, until all the air has gone. It is best to witness this if you can.

Pumping should be done slowly (using a low capacity reciprocating pump) so as not to impose dynamic pressure stresses on the vessel.

Test pressure is stated in BS 5500, ASME VIII or the relevant standard. This will not overstress the vessel (unless it is a very special design case). If in doubt use 150% design pressure.

Isolate the pump and hold the pressure for a minimum of 30 minutes.

What to look for

Leaks. These can take time to develop. Check particularly around seams and nozzle welds. Dry off any condensation with a compressed air-line; it is possible to miss small leaks if you do not do this. Leaks normally occur from cracks or areas of porosity.

Watch the gauges for pressure drop. Any visible drop is unacceptable.

Check for distortion of flange-faces etc. by taking careful measurements. You are unlikely to be able to measure any general strain of the vessel - it is too small.

If in any doubt, ask for the test to be repeated. It will not do any harm.

The vessel interior

It is important to make a thorough inspection of the inside of the vessel. This cannot be done properly by just looking through the manhole door you have to climb inside with a good light to be able to make an effective inspection. Check these points:

Head-to-shell alignment. Most manufacturers take care to align carefully the inside edges of the head-to-shell circumferential joint. Check that this is the case and that there is a nice even weld-cap all the way around the seam.

Nozzle 'sets'. Check the 'set-through' lengths of those nozzles protruding through into the vessel. Again, you should use the

approved design drawing.

Weld seams. Do the same type of visual inspection on the inside weld seams as you did on the outside. Make sure that any weld spatter has been removed from around the weld area.

Corrosion. Check all inside surfaces for general corrosion. Light surface staining may be caused by the hydrostatic test water and is not a cause for concern. If the vessel specification does not call for internal shot blasting, such staining should be removed by wire brushing. In general there should be no evidence of mill-scale on the inside surfaces - if there is, it suggests the plates have not been properly shotblasted before fabrication.

Internal fittings. Check that these are all correct and match the drawing. In many vessels, internal fittings such as steam separators, feed baffles, and surge plates are removable, with bolt threads being protected by blind nuts. The location of internal fittings is also important - make sure they are in the correct place with respect to the 'handing' of the vessel. It is also worth checking the fit of the manhole door and any inspection covers.

COMPRESSORS

Compressors work on simple principles but their tests are not always so easy to understand. Compressor designs vary from those providing low pressure delivery of a few bars up to very high pressure applications of 300 bars. For general industrial use the process fluid is frequently air, whilst for some specialized process plant applications it may be gas or vapour. There are several basic compressor types, the main difference being the way in which the fluid is compressed. These are: Reciprocating compressors. This is the most common positive displacement type for low pressure service air. A special type with

oil free delivery is used for instrument air and similar critical applications.

Screw compressors. A high speed precision design used for high volumes and pressures and accurate variable delivery.

Rotary compressors. High volume, lower pressure applications. These are of the dynamic displacement type and consist of rotors with vanes or meshing elements operating in a casing.

Other designs are: lobe-type (Roots blowers), low pressure exhausters, vacuum pumps and various types of low pressure fans. These are covered by different standards and acceptance procedures. Be careful not to confuse the types.

separate to the mechanism class. There is no imposed direct dependency between the two.

Air-drying paints

These are the most common type that you will meet in general engineering applications. They account for maybe 70-80 percent of all painting on indoor structures and equipment, or on outdoor installations which are not subject to highly corrosive atmospheres.

There are three main types:

Alkyd resins are used in primers and undercoats. Dry film thicknesses (dft) are generally 35-50 mm per coat. Expect to see cosmetic problems caused by the short term wet edge time. Adhesion is not a common problem if preparation is done properly.

Epoxy esters. Expect to find these on structural steelwork and storage tanks. They have better chemical resistance than all alkyd types. The inspection requirements are the same. You may need to guard against problems caused by poor spraying technique with this type of paint. Chlorinated rubbers. These are used for exterior protection of structural steelwork and fabrications in particularly corrosive

environments, such as coastal or offshore locations. The top coat is normally applied over a similar chlorinated rubber alkyd or zinc-rich primer, other types may result in adhesion problems. Coat thicknesses vary, but single coat dft is normally quite thin, about

50-60 (mm per coat due to the high level of solvent in the paint. Often, however, specifications will call for a single thick coat of 300400 mm. For this a thickening agent has to be added to the paint. Chlorinated rubber paints do not often suffer from intercoat adhesion problems, as subsequent coats fuse together quite effectively. They do suffer from problems such as sagging and peeling if the coating is applied too thickly.

Two-pack paints

Two-pack paints comprise a pigmented resin and a catalyst (or hardener) which are mixed together. They harden via a chemical

reaction rather than by evaporation of the solvent. The most common range is the 'epoxy' type which is highly resistant to water,

environmental, and chemical attack. Expect to see it specified anywhere where there are alkaline or acidic conditions, rather than for general service use. Note four specific inspection-related points:

Epoxy paints are generally 'hi-build' - an average dft is 100 mm per coat.

They have a very limited pot life once mixed, so application problems do occur.

The ambient temperature during painting is important. Epoxy will only cure successfully at temperatures above 7-8 C.

On balance there are more critical process-related factors than for air-drying paints. This means that monitoring of the application

process is an important part of the inspection process. The other type of paint in this category is two-pack polyurethane. This is also used for chemical-resistant applications. Dry film thickness can vary from 40-100 mm per coat, depending on the paint formulation used.

Primers

For metal which has been shotblasted it is common for a thin (30 mm) coat of zinc-based blast primer to be applied. This can be an 'etch' type - containing phosphoric acid to etch the surface - or a two-pack epoxy rich in metallic zinc. Zinc primers have inhibitive properties, the zinc providing cathodic protection to the iron. From an inspection viewpoint, a key issue is to ensure that these primers are applied immediately after shotblasting. The zinc must make intimate contact with the metal surface and not be restricted by corrosion products, which can form very quickly on a freshly blasted surface. For less critical fabrications, you may find zinc chromate or zinc phosphate (or occasionally red lead) primers being used. These are traditionally applied in quite thick coats and tend to be more common for on-site repair work than for new factory-built equipment. Their performance on steel that has not been fully cleaned is quite poor.

Preparation

Incorrect surface preparation is the most common root cause of failure of paint films, particularly those applied over common ferrous materials such as low carbon steel. Poor preparation does not always result in instant failure - a period of two or three years may elapse before the real problems become apparent. By then, however, the breakdown will likely be almost complete - the paint system having effectively given up its protection of the underlying metal. Proper preparation comprises two objectives. Removal of existing rust cells is one, but the main one is to eliminate any active (or latent) corrosion cells on the surface of the metal by removing the millscale. Millscale is hard brittle oxide formed during the steel rolling process. It causes 'mechanical' problems to a paint film because it expands and cracks, causing the paint to flake off. It also causes electrochemical problems because it is cathodic to steel, so it encourages rapid anodic attack on small areas of unprotected surface.

Three key points about millscale:

Proper preparation means eliminating all the millscale before painting.

Millscale cannot be removed by wire brushing, it is too hard. The practice of 'weathering' material only reduces the amount of millscale - it does not remove it totally.

The only way to remove millscale properly is by shotblasting with the correct heavy grade of grit. The degree to which a material is shotblasted is important. Grades of cleanliness of the blasted surface are covered by several well accepted standards (see later). From an FFP viewpoint, however, it is wise to be wary of grades of finish that do not specify complete removal of the millscale. I have shown these grades in Fig. 14.2 along with some key inspection points that you should check.

Application

Most fabrications, vessels, and larger equipment items that you meet during works inspections will have their paint applied by spray. Methods such as dipping (for small components), hand brush or roller application, and electrodeposition are less common. The technique of spray painting is one which relies heavily on the skill of the operator it is perfectly feasible for a properly prepared surface, with a well chosen paint system, to give poor results because of a poor application technique. It is not possible to describe the techniques of paint application here. References (1) are available. It is important, however, that as an inspector you have an appreciation of the main variables that have an influence on application. You will find them addressed in the painting procedures and record sheets used by good painting contractors. Use them also as a rough checklist of items to look at when you are witnessing paint application. The main variables are:

Spraying air quality. It should be moisture free, by using filter/dryers, in order to avoid contamination of the paint. Some techniques use an airless spray in which the paint atomizes due to pressure drop only as it exits the spray gun.

Paint mix and consistency. This is particularly important for twopack paint types. Shelf life and pot life restrictions need to be

complied with. Note that some paints are heated before being applied.

Ambient conditions. This is a key variable. Ambient temperature and relative humidity must be within prescribed limits (paint specifications clearly state what these are). High humidity (above 80 percent) and low temperatures (below about 4 C) are cause for concern. Dust-laden or salty conditions are also undesirable.

Painting defects

The main painting defects that you are likely to see are described below.

Sagging and curtaining

These are easy to identify, consisting of obvious areas of paint overthickness where the wet film has sagged under its own weight.

You will most often see it in wide horizontal bands on large vertical surfaces. It is caused by applying too much paint. In extreme cases it will develop into 'runs' down vertical surfaces. Although unsightly, it is not necessarily cause for rejection, unless it is widespread or is accompanied by surface preparation problems.

Orange peel effect

In this case, the paint surface when dry resembles orange peel, with a dappled surface, sometimes accompanied by small blisters. This may be a problem of paint consistency or poor spraying technique. In most cases the adhesion of the film is affected, so significant areas of orange peel effect are a clear cause for rejection.

Wrinkling or lifting

At first glance this appears similar to the orange peel effect with a blistered appearance to the film surface. The quickest way to

differentiate is to run your hand over the surface to check whether any significant lack of adhesion is present. If the surface film is clearly 'loose' (especially at the edges of any large blisters) then this is indicative of an intercoat problem. The most serious type is caused by incompatibility of the solvents used in successive paint coats, i.e. a paint system error. A similar, but less pronounced, effect can be caused by the operator not allowing the correct time interval between coats.

Rough surface finish

A poor surface finish is normally indicated by the loss of gloss on the top coat. It is easiest to see on large flat surfaces. The main causes are condensation or airborne dust and dirt. Unless dirt contamination is very serious, it should be possible to rub down and recoat, without having to remove the existing paint coats.

Pinholing

This has the appearance of large numbers of small concentrated pinholes in the paint surface. The most likely cause is contamination of the paint by oil or water. Extensive pinholing does require a full repaint, otherwise the FFP of the paint system is likely to be affected.

Thin areas

This is the most common defect that you will encounter - expect to find components with overall too-thin coatings as well as those where only individual areas such as edges have insufficient application. Provided that the primer and undercoat films have been applied, then the top coat thickness can easily be built up by further application of paint. If (as occasionally happens) the primer or undercoat are much too thin, or missing, then the coating will not meet its FFP requirements. Strictly, an additional thickness of top coat is not an acceptable remedy for a deficient primer layer, as primer often has an inhibitive role whereas the main function of the top coat is to protect against weathering and mechanical damage.

Dry film thickness (dft)

The dft of a paint coating is an important parameter. Assuming correct preparation and application, the durability and protective properties of a coating are related directly to its dft. Adequate thickness is necessary in order for the film to have sufficient electrolytic resistance to prevent the formation of the local galvanic cells that cause corrosion. Most painting specifications show the dft required for the separate primer, undercoat and top coat, as well as for the three together. You will see that these are shown as minimum dft. Sometimes a maximum is also quoted but often there is none - it is inferred that a thicker dft is acceptable, as long as this is not so thick as to result in peeling, sagging, or other defects associated with excessive application. It is only possible to obtain a true dft reading when a paint film is completely dry - this can be up to 24 hours after application of the last coat, depending on the type of paint. During application the painting operator checks wet-film thickness using a 'comb' or 'wheel' gauge. This thickness then reduces to the dft as the solvent evaporates. It is difficult to quote general relationships between wet and dry film thickness as it varies with the type of paint (it will be shown on the paint manufacturer's data sheet if you need to know it when you are witnessing paint being applied). Simple hand-held meters are used to measure dft. They work on electromagnetic or eddy current principles and provide a direct digital readout. Some have a recording and print-out facility. Here are a few general guidelines to follow when witnessing dft measurements:

Calibration. Do a quick calibration check on the dft meter before use. Small test pieces, incorporating two or three film thicknesses, are normally kept with the meter. You can also do a quick zero calibration test on a convenient exposed steel surface - a machined surface is best.

Test areas. Test a large number of points, two or three are not sufficient. Make sure you include vertical and horizontal areas,

corners, radii, and edges. Pay particular attention to enclosed areas. Use your appreciation of which areas would be difficult for the

operator to spray easily.

Results. Specified dft requirements are generally accepted as being considered 'nominal' or 'average' values. As a rule, the average dft measured (from multiple readings) over a minimum 1 m2 area should be greater than or equal to the specified level but, there should be no individual dft readings of less than 75 percent of the specified level. This means that there is always room for some interpretation, unless a purchasing specification is very specific about the number and location of measurements that are to be taken. If in doubt, the best standard to look at is ISO 2808 (BS 3900 part C5).

Paintwork repairs

The easiest place to do repairs to paintwork is in the works, before shipping the equipment to the construction site. Some construction sites are well equipped and have the necessary skilled subcontractors for painting large items, but many are not. The site may also have environmental problems such as high humidity, dusty or salty air. Repair techniques depend on whether the faults you have found constitute a major FFP non-conformance (wrong preparation, system or application) or whether they are cosmetic. A situation involving a major non-conformance normally has only one solution - remove the faulty paintwork and repeat the process, taking care to eliminate the previous faults. Given that serious faults have already occurred there are several points that you should consider.

Do not start the repair without a proper diagnosis of the original fault and what caused it.

The only real way to obtain proper preparation is to shotblast completely to a minimum grade of Sa2 1/2 Witness this activity -

repairs are often done to emergency timescales which do not encourage careful and thorough surface preparation. You need to check the new paint system. The best way to do this without wasting time is to talk directly to the paint manufacturers - they can give you a very quick response about the suitability of a paint system for a particular purpose, compatibility of coats, and details of preparation and application techniques required. Witness the application of all three coats if this is practical. If not, make sure that reliable visual and dft checks are done after each coat. Repeat all the final checks again after the repainting is finished. This is one area where site rectification of any subsequent defects would not be good practice. Expect some equipment manufacturers to place you under pressure to release the item before you have checked it properly. Decline such invitations.

Cosmetic painting defects can be treated differently. Unless the substrate material is actually exposed, or you can find evidence of a

more serious problem, it is up to you whether or not you feel it necessary to witness or reinspect cosmetic repairs. Cosmetic defects should be repaired even if you have concluded that they do not compromise fitness for purpose. Try not to say or do anything that would enable these defects to be misunderstood out of context by a distant client or end user. Do not base a whole inspection report around lists of cosmetic defects - not if you want to be taken seriously.