A look back at the development of Water Leak detection in Computer & Server Rooms

Water leak detection came about in the late 70’s when computer rooms were in there infancy. Computer rooms as today, contain air conditioning which contains water for humidifiers and sometimes chilled water for cooling. Because of the large amount of power cables, data cables and water pipes needed in the room, a raised floor was and is still used to hide all services. Unfortunately, any water leaking under this raised floor would not be discovered until power/data connections were immersed in water and the computer stopped working.

Up to the mid 80’s water was detected using spot probe sensors. These units would consist of either an etched PCB or two metal electrodes. Sensing for water was done using a DC voltage in one sensor whilst looking for a return signal in the other. Providing no return signal was seen in the return sensor, no water was present. The problem using this type of system was erosion of the sensors due to electrolyses and the limited area of water detection, water could flow away from the sensors and not be detected until too late.

During the mid 80’s water detection cable was developed. The advantage of this type of sensing being that water is detected along the entire length of cable. This allowed areas or equipment containing water to be surrounded insuring that leaks were found no mater what direction the water flowed.

From the mid 80’s through to today advances have mainly been with the alarm panels and the reporting of water leaks. Today you can be Texted, Emailed, receive a phone call, record it on a building management system or just have the simple buzzer and lamp.

My involvement in water detection by Richard Benford

I was first asked to design a water detection system in the late 70’s whilst working as for Vikingshaw Products Ltd. Our Mother company Vikingshaw Ltd at that time built computer rooms around the country and Vinkshaw Products supplied them with Power Distribution Units etc. The first systems were simple in there design being DC based with PCBS for sensors and control units with a simple buzzer and lamp. Unfortunately it wasn’t long before I discovered that DC was not the correct way to detect water as our sensor copper tracks would disappear if left in water for a few hours. It was from this point that I used an AC signal in the sensors to stop them eroding away. In the Early 80’s I went into partnership and started a company called Wayscale Ltd. During the 80’s, 90’s I developed a process to manufacture water detection cable and alarm systems to display water leaks from one to 128 different areas or zones. The top of the range multi zone control unit used addressable outstations with four independent water detection zones and a 24 alpha numeric display to advise of the water leak location in words and numbers. In 2003 and by mutual consent both my partner and myself decided to stop trading and closed down the factory. As part of the breakup of Wayscale, I took with me the product design rights including the manufacture of the water detection cable and started CMR electrical Ltd with two of my sons. From 2003 to the present, developments have mainly been with the alarm controller due to the improvements in electronics. During the past three decades I have been responsible for the design of thousands of water detection systems which have been installed in every application imaginable from large government security buildings through to small server rooms in the UK and abroad.

Considerations when designing a water detection system

A number of factors need to be taken into account when designing a water diction system. Failure to do so could lead to systems not detecting water when required to do so.

1. The uses of an alternating current (AC) not direct current (DC) in the sensors. Direct Current will erode the sensors if left in water for long periods leaving the system unusable for any subsequent alarms.

2. Sensitivity adjustment to allow damp areas or condensed water from AC units to be ignored, but still allow large leaks to be detected.

3. Zone cross talk leading to false alarms or wrong location caused by one signal from one zone using Earth (floor jacks, conduit etc.) to interfere with other zone causing one or both zones to go into alarm without the presents of any water.

4. Quick sensor recovery after a water leak. Detection cables and spot probes should be capable of being removed and wiped with a cloth or tissue paper to remove water allowing the zone to quickly reset.

5. Sensitivity of the sensors, electrodes too close to each other in both water detection cable and spot probes will cause false alarms due to condensation or water droplets.

Using an alternating current stops electrolyses which causes the sensors to disintegrate and the system unable to detect water. The use of alternating current also allows the sensors to continually monitor for water even when submersed in water. This allows the system to quickly automatically self reset once water has been removed for the sensor with no further action by an operator to reinstate the system back to normal operation.

Where do our systems go

We have supplied systems to look for water leaks in computer rooms, server rooms, plant rooms, UPS rooms, office shower rooms, office tea areas, vending machines, high level metal beams in ice rinks, high level water pipes, under floor office heating pipes, cesspit overflow, under water tanks, Living walls, drip trays in and under air conditioning units and bunded areas for tanks. Our customers range from one of the largest online betting company’s, one of the largest and well known server room hoisting centre in London, Jersey Telecom, numerous large office blocks in London and numerous small server rooms around the UK and Ireland.