Glass Fused to Steel and Bolted Tanks

Glass Fused to Steel and Bolted / Sectional tanks have been used extensively in the water, waste water and biogas industries due to their cost effectiveness and speed of erection when compared to other tanks. Most commonly the steel panels are protected using glass enamel, epoxy or galvanizing. All of these treatments have a finite lifespan, which when reached or when the surface protection leads to corrosion and in aggressive environments perforations.

In terms of preparation these tanks are very similar to Steel Tanks, however there are some intricacies that must be taken to into account.

Glass fused steel tanks are a challenging tank lining application with intricacies that many others overlook and an area that we have researched extensively.

Glass fused steel is marketed as an extremely durable solution, however whilst the finished surface offers good chemical resistance, this does not tell the full story.

As can be seen in the micrograph the chemical resistant enamel layer makes up only a very small percentage of the total thickness. Below the dense enamel surface is much more porous matrix.

micrograph showing the chemical resistant enamel layer makes up only a very small percentage of the total thickness

Once the thin enamel surface layer is broken through mechanical damage, moisture and aggressive chemicals are absorbed into the inherently porous substrate. The presence of moisture against the steel creates a corrosion cell, which if accelerated by aggressive agents such as biogenic sulphuric acid, rapidly leads to perforation, as seen.

Taken from a research paper titled ‘Degradation of Vitreous Enamel Coatings’ by A Conde and JJ de Damborena;

As part of our own research four samples were sent for analysis by scanning electron microscope as shown below;

Glass Fused Steel

Grit Blasted Glass Fused Steel

Grit Blasted Steel

SPI Corrolast HT

When viewed under an electron microscope at 100 times magnification, defects can clearly be seen in the surface of the enamel, as highlighted in yellow. These may not be picked up with a holiday/continuity test if they do not penetrate to the steel substrate. These would provide a pathway for the creation of a corrosion cell.

Gfs 100x - Glass Enamel - Shiny dark green sample

When the blasted glass fused steel is steel is examined at 500 times magnifcation as shown, the porosity can clearly be observed as previously mentioned. It is this porosity that provides a pathway for mositure to reach the steel substrate beneath.

This phenomenon can be seen even by the naked eye by simply applying water to a grit blasted glass fused steel plate.

As can be seen, the enamel is no longer waterproof, and even after removing excess water with a cloth, a water mark persists.

When re-lining sectional glass fused tanks there are some important factors to take into consideration. These are as follows:

Adhesion – Even prepared glass fused steel is a difficult surface to adhere to.

Lateral Movement – The bolted modular nature of these tanks means that some lateral movement between panels is expected when emptying and filling.

Deflection – Within the panels themselves deflection is expected.

These points are addressed below in more detail.

As can be seen, the enamel is no longer waterproof, and even after removing excess water with a cloth, a water mark persists.

Adhesion

Nearly all tank linings/protective coatings gain adhesion to a surface through adhering to surface profile/roughness. In a liquid state the coating flows into the imperfections in the surface and once cured locks into positions.

Glass fused steel is an extremely dense surface and in an unprepared state virtually free from any surface profile/roughness for a coating/lining to adhere to. The widely recognised method of raising a surface profile is abrasive blasting. However not all blasted surfaces are created equal!

Referring back to the samples sent for analysis by electron microscope, these were also subjected to optical-interference profilometry. This creates a numerical and visual representation of the surface roughness – useful for understanding how coatings will adhere to a particular surface. Grit blasted steel is widely recognised as one of the best surfaces for coatings to adhere to.

This is shown for both blasted glass fused steel and blasted steel as shown in the image.

What is interesting but also very important to understand when you compare the two, is that the blasted glass fused steel has a greater profile range (RA), something that could be measured with site based equipment such as surface profile needle gauges. However when you look at the visualisation of the the type of profile it is clear that the blasted glass fused steel is much more rounded, without the peaks and troughs that blasted steel offers and what makes it such a good surface for coatings to adhere and ‘anchor’ to.

This can be seen visually when seen under a scanning electron microscope at 100 times magnification as shown on the next page.

Clearly visible is the relatively smooth surface of the blasted glass fused steel, far more difficult to gain adhesion to with a coating than blasted steel.

This smoother surface means that in order to obtain good adhesion to glass fused steel a coating must bond both chemically and mechanically, as mechanical adhesion alone is not sufficient.

To prove the effectiveness of the adhesion of our proposed lining (specifically developed by ourselves for application onto glass enamel), several adhesion pull of tests using different generic lining technologies were conducted.

The first material tested was our own epoxy primer (not being proposed for this application). This was applied to blasted glass fused steel, with a surface roughness/profile in excess of 100 microns when measured with a surface profile needle gauge.

As can be seen in the video, the adhesion failed at 6N/mm².

As important as the value at which adhesion fails, is the mode of failure. Here you can see that at the point of failure the coating has removed the top surface of the glass fused steel.

The next material tested was a Lloyds approved GRP Lining system – commonly known as fibreglass. This was applied to blasted glass fused steel, with a surface roughness/profile in excess of 100 microns when measured with a surface profile needle gauge.

Again this material failed at 6N/mm² as shown in the video;

Note the much reduced interaction with the glass fused steel as evident by the reduced presence of the material on the back of the test dolly. This suggests poor penetration of the substate and the pretence of adhesion is being given by the mechanical properties of the material.

The third material tested was Corrolast DSP – a primer that we specifically developed for adhering to glass fused steel. As with SPI Corrolastic HT this forms a chemical bond with the substate, without relying on surface profile at all. To prove this, the material was applied to smooth glass fused steel, without any surface profile present.

As shown in the video below, this adhesion test failed at 9N/mm2 – a 50% improvement compared with the other two materials.

Why Does Adhesion Matter?

Put simply, the adhesion of a lining is fundamental to its performance. For example, if mechanical damage were to occur, a poorly adhered lining that has not bonded tenaciously to the substate will rapidly fail as water and chemical agents will separate it from the substrate.

Whilst it would be possible for all linings to bond well with the steel beneath, this would require the removal of all of the glass fused steel. Not only does this have increased environmental implications both in terms of CO2 emissions and greater waste creation, this also represents poor value for the client. The reason for this is that they would effectively be paying to remove material that would contribute towards improved corrosion protection through greater total film thickness.

In this application also, due to the potential for movement and deflection in the tank which we will cover later, excellent adhesion is essential in order to prevent stressed areas of the lining from debonding. If an area of the lining were to debond, this would create an interstitial space behind the lining. Through osmosis (as all organic coatings are semi permeable membranes to varying degrees), moisture could then form, creating the conditions required for corrosion to occur unchecked beneath the lining.

Lateral Movement

Due to the bolted nature of these tanks there is potential for lateral movement, both after an initial fill and, during emptying and filling, up to approximately 1mm of movement over the lateral joint.

Whilst 1mm of lateral movement perhaps does not sound significant, as the starting point across the panel joint is effectively 0mm, any movement from this point is effectively hundreds of times elongation.

Brittle coatings such as conventional epoxy and GRP (Fibreglass) do not have the elongation properties to accommodate this movement when present.

Our proposed lining has been tested in accordance with BS EN 1062-7 to determine its crack bridging ability.

As can be seen in the below image from the test, the material reached in excess of 15mm elongation from 0mm before splitting. This means that the material has more than sufficient elongation properties to accommodate what could be expected in tanks of this construction.

This is in sharp contrast to epoxy, vinyl ester and GRP lining systems which typically have little to no elongation and flexibility and are likely to crack and fail when a tank is refilled after completion of the lining application and in operations of emptying and filling thereafter.

Deflection

As well as the lateral movement that can exist within this type of tank, the presence of deflection forces within the panels themselves also needs to be considered. Any lining material needs to be sufficiently well adhered to remain attached, but also flexible enough so as not to crack or fatigue from the deflection forces. Below is an illustration of the deflection that can be seen in the vertical plane

The largest tank manufacturer in the world CST Industries, tests the flexibility of their factory coated epoxy panels in accordance with ASTM D522 – as per the attached data sheet. The factory applied epoxy coating passes a 1/8th mandrel, without cracking, splitting or debonding.

We therefore subjected some industry standard materials to this same test standard in order to demonstrate that our proposed lining has the same flexibility and adhesion as a factory applied coating, meaning that it will offer the same durability. The tests can be seen in the video.

SPI Corrolastic HT Bend Test – PASS

The largest tank manufacturer in the world CST Industries, tests the flexibility of their factory coated epoxy panels in accordance with ASTM D522 – as per the attached data sheet. The factory applied epoxy coating passes a 1/8th mandrel, without cracking, splitting or debonding.

We therefore subjected some industry standard materials to this same test standard in order to demonstrate that our proposed lining has the same flexibility and adhesion as a factory applied coating, meaning that it will offer the same durability. The tests can be seen in the video.

SPI Corrolastic HT Bend Test – PASS

Convetional Epoxy Bend Test – FAIL

GRP/Fibreglass Bend Test – FAIL

Another important difference when preparing these tanks is that there is often an excess of sealant/mastic between panel joints and around bolt heads. This must be removed as this is often compromised hiding corrosion beneath which if left unchecked will continue and lead to perforations.

Where perforations exist these can be repaired prior to the tank lining being applied by bonding or mechanically fixing plates of the same thickness and material.

As with all of our tank relining projects the same quality assurance steps would be taken and implemented by our own in house ICORR accredited paint inspectors. Once cured our tank linings are tested for pin holing and porosity using a DC holiday spark tester as a pin hole is a point of failure where moisture and chemicals can reach the steel substrate.