Diagnosis
Why wraparound sleeve repairs fail in the ground
Six failures account for most of them. Five are settled before the torch is lit.
Every headline peel figure is a 25 °C figure
The adhesive is the repair. The backing holds shape and takes the shovel. The hot melt is the only thing keeping water off the metal, so the number that decides the job is peel strength, and it is quoted twice.
On the three-layer pipeline sleeve manufactured by Gala Thermo Shrink Pvt. Ltd., peel to steel is 100 N/cm at 25 °C. The same bond is 30 N/cm at 50 °C. The method is ASTM D2671. That is a 70 percent fall across 25 degrees, printed in the same cell as the figure everyone quotes.
Nothing about 50 °C is exotic. A buried line carrying warm product sits above it. So does an open trench in Gujarat in May, and so does an exposed span in afternoon sun. The joint does not let go at the moment it warms up. It holds at 30 N/cm until something pulls, and then it goes.
Above that pair there is one more figure and then silence. The hot-melt softening point is 95 °C. Between 50 °C and 95 °C nothing on the sheet tells you what the bond is worth, because nothing in that band was measured.
Peel is not published for the cable repair codes at all
The table below belongs to the GTWS three-layer pipeline sleeve. The wrap-around sleeve sheet covering GWS and GMRS carries no peel strength, no adhesive softening point and no recovery temperature. Do not carry a pipeline figure across to a cable job. Ask the manufacturer for peel against the jacket you actually have, PVC, XLPE or PE, and for the temperature it was measured at.
Hot-melt peel strength, cold and hot
| Bond | Peel at 25 °C | Peel at 50 °C | Method |
|---|---|---|---|
| PE / steel | 100 N/cm | 30 N/cm | ASTM D2671 |
| PE / FBE | 120 N/cm | 35 N/cm | ASTM D2671 |
| PE / PE | 110 N/cm | 30 N/cm | ASTM D2671 |
As published for the GTWS three-layer sleeve on the manufacturer's product page for it. No separate GTWS datasheet exists in the published library, so this page and that one are the whole record.
Six ways a wraparound repair is lost
Each leaves a different mark.
1. Heat in service
You sized the repair on the cold column. The joint works in the hot one. Take the highest metal temperature the asset reaches, not the ambient on the day of installation, and read the peel row against that.
2. Not enough overlap
The seam is the one band of sleeve with no asset behind it. Minimum overlap steps with code on the GMRS chart: 25 mm up to GMRS-70, 50 mm at GMRS-115 and GMRS-160, 75 mm from GMRS-180 upward. Cut a sleeve to the circumference and the whole margin is gone.
3. A substrate nobody prepared
Peel is a coupon result and the sheet never says how the coupon was prepared. Neither wraparound sheet carries a preparation step. The manufacturer does print one, on the pole cap sheet: clean and de-grease the surface that will meet the sealant. Nothing published carries that instruction across to a cable jacket or a welded cutback, so agree it in writing and record what the crew did.
4. Too much torch
Longitudinal change is minus 10 percent maximum, published on the post-tensioning coupler and the U-bolt coating sheets to ASTM D2671. Recovery keeps taking length while heat goes in, so a 500 mm sleeve can finish near 450 mm and walk its own ends off the cutback. The pole cap sheet gives the torch rule the sleeve sheets omit: a soft blue flame with a yellow tip, no pencil flame, kept moving.
5. The wrong outer diameter, either way
Too large will not grip. Too small will not close. And the published band is not always one band: for GWS-105/28 the manufacturer's product page gives 85 to 35 mm and its printed sheet gives 92 to 30 mm. Work to the narrower of the two, and measure the asset rather than trusting the drawing.
6. Cathodic disbondment
On a protected line, the current that keeps the steel safe is the same current that lifts a sound coating outward from a defect. GTWS is quoted at 3 rad to ISO 21809-3. This is the one entry on the list that is not the crew's fault.
What each of these looks like when you dig it up
This is a reading key built backwards from published properties. It is not a failure database, and no field-failure record for these parts is published anywhere. Treat it as where to look first, and photograph the joint before anyone disturbs it.
- Sleeve whole, adhesive glossy, and the whole assembly slides on the asset. The backing did its job and the bond gave up. The question is what the joint was running at, not what the part is rated at.
- A bare band showing at one or both ends. Axial recovery has pulled the sleeve short of the cutback. Compare the finished length against the flat length on the chart before anyone blames the adhesive.
- The seam has opened and everything else is still bonded. Overlap was short, or somebody trimmed the sleeve to the circumference. Measure what is left of the tongue and read it against the minimum for that code.
- Adhesive lifted in a ring around one defect, sound in every direction beyond it. That geometry is cathodic disbondment rather than workmanship. Measure the radius outward from the holiday and set it against the 3 rad figure.
- Mill scale, grease or a bright shine on the metal under the released adhesive. The bond never wetted out. No published peel figure applies to that surface, because nothing was measured on it.
- No bead of adhesive squeezed out at either end. The pole cap instructions treat a visible bead as the mark of a finished shrink. Its absence on a recovered sleeve says the hot melt never flowed.
- The sleeve is displaced, wrinkled or dragged along one face. The same instructions say to let the part cool before any mechanical strain reaches it. Backfill dropped onto a warm joint is mechanical strain.
- The colour-change paint turned and the joint failed anyway. That paint is an outer-wall thermometer. It cannot see the adhesive or the seam, and both of them are underneath it.
The seam is where an excavated failure starts
The stainless channel in the manufacturer's photograph holds the two edges together while the sleeve recovers. It is not the seal. The seal is the hot melt inside the overlap beneath it, and that strip is the only place in the whole repair where adhesive bonds to adhesive instead of to the asset.
Design geometry and failure geometry are not the same drawing. On the chart the overlap is a minimum of 25, 50 or 75 mm depending on the code. In the ground it is whatever survived the sleeve being trimmed on site, the asset measuring larger than the drawing said, and the seam being closed while the joint was still cold.
The pipeline sleeve settles the argument by making the closure a separate part. Every GTWS code ships with its own patch, 100 mm long against the smallest code and 200 mm against the largest, at a constant 500 mm width. There the overlap has a code and a dimension instead of being a length somebody estimated in the trench.
Where each of these is settled
- Heat shrink sleeve size chartGWS and GMRS bands, recovered walls and the overlap column.
- Pipeline sleeve size chart, 159 to 1016 mmFlat length against pipe outer diameter, with the closure patch supplied against each code.
- Cathodic disbondmentWhat 3 rad is measured against, and the seven conditions the published row leaves out.
- How to install a wraparound heat shrink sleeveThe sequence, and where published instruction stops and judgement starts.
- Wraparound or tubularA seam you never needed is a seam that cannot open.
- Reading a heat shrink datasheet without taking it on trustWhat ASTM D2671, ASTM D638 and ASTM E28 each measure.
Send the failure, not the part number
Give the substrate, the highest temperature the joint sees in service, the measured outer diameter, and what the sleeve looked like coming out of the ground. That is enough to say whether the part was wrong or the installation was. This site answers the technical half; anything commercial belongs to the manufacturer.