If a metal structure such as a gas pipe corrodes excessively, gas could leak out through the corrosion and cause an explosion. Thus it is important to monitor the condition of gas pipes. However, it is a burdensome task to monitor the condition of pipes, especially those that form part of a large network. It is made even more burdensome if the pipe is coated, for example, with insulation as it is time consuming to remove the insulation, inspect the pipe and replace the insulation.

According to the present invention, there is provided an apparatus to predict the risk of or potential for corrosion of a metal structure, the apparatus comprising:- a temperature detector; a fluid detector; and a processor to predict the risk of or potential for corrosion based on information received from the detectors.

Such an apparatus may be installed on a metal structure such as a gas pipe for example to monitor the potential for corrosion of the pipe. If the pipe is surrounded by insulation, a portion of the detectors may be located under the insulation so that an analysis may be performed without having to remove the insulation.

The temperature detector preferably includes an optical fibre which may be provided along at least a portion of the structure or pipe. The fibre may run axially along the outside surface of a pipe or may be wound, for example, spirally around the outside surface of a pipe. Such optical fibre based temperature sensors are commercially available and are well known to those skilled in the art. The use of a temperature sensor incorporating an optical fibre may enable the temperature at different locations on the metal structure or pipe to be determined which may enable the location of any potentially corrodable locations to be identified.

The fluid detector preferably includes an optical fibre which may be provided along at least a portion of the structure or pipe. The fibre may run axially along the outside surface of a pipe or may be wound, for example, spirally around the outside surface of a pipe. Such optical fibre based fluid sensors are commercially available and are well known to those skilled in the art. The use of a fluid sensor incorporating an optical fibre may enable the detection of fluid at different locations on the metal structure or pipe to be determined which may enable the location of any potentially c orrodable locations to be identified.

The temperature and fluid detectors could be combined into a single optical fibre detector, or into a single cable comprising more than one fibre.

The processor may be a microprocessor or computer, for example. The processor

receives and monitors outputs from the detectors and when appropriate temperatures and the presence of fluid is detected indicates the risk of or potential for corrosion based on the detected temperature, the presence or amount of fluid and preferably also prior knowledge of the material. The potential for corrosion may be indicated by the illumination of a light, sounding of an alarm or an indication on a computer screen, for example.

If the temperature and fluid detectors are able to identify the location of the parameter that they detect, the processor may identify the risk of or potential for corrosion at different locations, for example, as a function of distance along the fibre (s).

According to a second aspect of the present invention, there is provided a method of predicting the risk of or potential for corrosion of a metal structure, the method comprising : - detecting the temperature of the metal structure; detecting for the presence of fluid on the metal structure; and using the results of the two detections to predict the risk of or potential for corrosion.

An example of the present invention will now be described with reference to the accompanying drawings in which:- Figure 1 shows an apparatus for predicting the risk of or potential for corrosion of a metal pipe;

Figures 2 and 3 illustrate the determination of susceptibility to corrosion for two pipes of different material; and Figure 4 shows an alternative arrangement of the detectors on the metal pipe.

Figure 1 shows a metal pipe 10 with a pair of optical fibres 11,12 wrapped around it.

One fibre 11 is used to measure temperature by measuring changes in the backscattered radiation as is well know in the art. The other fibre 12 is wound around a water swellable element 13 so that when the element 13 swells, it causes the fibre 12 to bend, the bending or bend losses again being measured by changes in the backscattered radiation.

Each fibre 11,12 is provided with a device 21,22 which both emits a light pulse into the r espective fibre a nd detects a ny b ackscattered radiation. S uch a d evice 2 1, 2 2 could be provided at both ends of each fibre to ensure that detection is still performed even if there is a breakage in one or both of the fibres.

The outputs from detection devices 21,22 are passed to a processor, in this case a personal computer 30. The computer displays a plot of temperature 31 as a function of distance along the pipe 10, a plot of the presence of fluid 32 as a function of distance along the pipe 10 and a plot of the risk of or potential for corrosion 33 along the pipe 10. However, the computer need only display the plot of the risk of or potential for corrosion 33. The computer 30 uses the data regarding temperature 31

and the presence of fluid 32 in an algorithm to determine the risk of or potential for corrosion 33. The computer 30 may also take account of prior knowledge of the material from which the pipe 10 is made in the determination of the risk of or potential for corrosion 33.

Examples illustrating the determination of susceptibility to corrosion under insulation (CUI) for two pipes of different material are described below with reference to Figure 2 and Figure 3.

The first example relates to an insulated austenitic (300 series) stainless steel pipe.

Corrosion in austenitic stainless steel is manifest as pitting or chloride stress corrosion cracking. Austenitic stainless steel pipes operated at temperatures between 60°C and 240°C are known to be susceptible to corrosion under insulation with most corrosion occurring when operated in the temperature range 60°C to 90°C. Figure 2a illustrates the mean temperature (°C) and per centage time that water is detected at points along the length of a 30m austenitic stainless steel pipe. The temperature detector shows gradually increasing temperatures from O°C to 40°C along the length of the pipe and the fluid detector shows intermittent detection of water adjacent to the pipe at various locations. Figure 2b shows the corrosion under insulation (CUI) likelihood at various points along the length of the pipe, with 1 being low likelihood, 2 being medium likelihood and 3 being high likelihood. The CUI likelihood at each point along the pipe is based on the temperature reading, per centage time that fluid is detected and

prior knowledge of the material from which the pipe being tested is made. As can be seen, the insulated austenitic stainless steel pipe with conditions illustrated in Figure 2a has a low CUI likelihood along its entire length.

The second example relates to an insulated carbon steel pipe. Corrosion in carbon steel is manifest as generalised or localised loss of wall thickness. Carbon steel pipes operated at temperatures between-4°C and 120°C are known to be susceptible to CUI, with most corrosion occurring when operated in the temperature range 60°C to 90°C. Figure 3a illustrates the mean temperature (°C) and per centage time that water is detected at points along the length of a 30m carbon steel pipe. The temperature detector shows operating temperatures in the range 80°C to 100°C and the fluid detector shows permanent detection of water adjacent to the pipe at fixed locations.

Figure 3b shows the CUI likelihood at various points along the length of the pipe.

The CUI likelihood at each location is based on the temperature reading, per centage time that fluid is detected and prior knowledge of the material from which the pipe being tested is made at each location. As can be seen from Figure 3b, the illustrated carbon steel pipe with conditions illustrated in Figure 3a has a high likelihood of CUI at two points along its length and a low likelihood of CUI on the remainder of its length.

Figure 4 shows an alternative arrangement of the fibres 11,12 arranged longitudinally along the length of the pipe 10. In this example, four pairs of fibres 11,12 are

arranged equiangularly around the circumference of the pipe. However, any number of pairs of fibres 11,12 may be provided along the pipe 10 and if more than one pair of fibres 11,12 is provided, they may be spaced by any desired amount.

The fibres 11,12 forming part of the temperature and fluid detectors may be arranged over a large network of pipes with one or a few processors to analyse the results and indicate when there is a risk of, susceptibility to or potential for corrosion in the network and the location of any high risk areas which can be dealt with appropriately.

This reduces or overcomes the need for periodic inspection of pipes for corrosion which is very labour intensive and thus expensive and can be unreliable if, for example, used with insulated pipes where the insulation is removed for inspection at only a number of test points. Furthermore, the use of constant or regular monitoring by the method of the present invention is likely to highlight the risk of corrosion before serious damage occurs enabling far less expensive repairs to be performed.

The invention is applicable to any suitable metal structure which is susceptible to corrosion such as tanks for storing fluids or processing plants, for example, in addition to pipes as illustrated in the examples.