Tony Turp

Control Systems Technical

Specialist

Barking Power Station

csy@thames-power.com

UK

Barking Power Station, owned by Barking Power Limited and operated by Thames Power Services, is one of the largest independently-owned generating plants in the UK, capable of generating 1000MW of electricity – about 2 % of the peak electricity demand in England and Wales. The station uses Combined Cycle Gas Turbine (CCGT) technology, with gas as its primary fuel, and is able to generate at around 50 % thermal efficiency while producing low levels of emissions.

The initial plant design was for “baseload” operation but current market conditions have resulted in running on a two-shift pattern, responding to predictable peak level demands for energy. Failure to supply power when required, known as ‘missing a window’, results in lost revenue, which can be in the region of hundreds of thousands of pounds, depending on the unit price per MW at the time. Being capable of providing consistent levels of production when called upon is therefore critical to maximising plant profitability.

To minimise the risk of being unable to produce power, it is important to manage critical plant assets, identify potential problems early and schedule required maintenance to help prevent unexpected downtime and increase plant reliability. We are constantly looking for ways of improving our ability to identify problems earlier, and to prevent them from becoming serious issues. 

Click Image to Enlarge

Wireless acoustic transmitters at Barking Power Station.

Leaking Valves and Steam Traps

Steam loss within the plant is a particular area of concern. Operators perform frequent rounds to identify leaks from valves or malfunctioning steam traps during normal operations, but the dynamic nature of the process makes this difficult. Large, single leaks will normally be easy to detect, but smaller leaks may go unidentified for 2–3 weeks. Although we lose low volumes of steam through these smaller leaks, multiple leaks can potentially add up to four tonnes of steam lost per hour before the problem is detected by deterioration in plant performance. Ideally therefore, we want to identify failed steam traps and leaks caused by malfunctioning valves as early as possible.

Online acoustic transmitters provide the perfect solution to this problem. Fitted to individual steam traps and valves, these non-intrusive devices use acoustic “listening” technology, combined with temperature measurement to identify turbulent flow generated by failed steam traps or leaks.

Historically, the difficulty with deploying these devices is that we have around 100 steam traps onsite and many more control valves. These are distributed across the full extent of the plant, sometimes in very remote areas. Traditionally, adding a new acoustic device has required supporting power and data cabling which in most cases is not available. Installing a new cable infrastructure is difficult, time consuming and cost prohibitive.

The solution lies in WirelessHART® networks and the availability of Rosemount 708 wireless acoustic transmitters from Emerson.

We were very familiar with the Smart Wireless technology, having previously installed a wireless network in the plant to help us gather stranded diagnostic data from our valves and to enable ad-hoc data collection using wireless transmitters as portable data-recorders. By adopting a plantwide wireless network for these functions, we were able to reduce costs and save time by removing the need to add additional wiring infrastructure. The wireless network has proven to be very reliable, despite the tough environment of the power plant with its metal structures and piping. We were therefore very confident of using our installed wireless network to include the monitoring of steam traps and valves.

The plant is split into two areas consisting of a 400 MW unit with two boilers and a 600 MW unit with three boilers. To cover both areas we installed two new wireless field networks that will support up to 100 devices on each wireless gateway.

The wireless acoustic transmitters were installed on selected steam traps, where they can pick up a change in noise level from the expected footprint, which may indicate a leak. As soon as this happens, the device transmits the information to our DCS via a WirelessHART network, so that maintenance can be alerted and repairs scheduled for the next period of no energy production. This reduces the risk of significant leaks taking place during periods of production. Early indications lead us to believe that analysing the data coming from the devices may also give us indication of the changing aspect of a problem as it occurs.

Having already installed the Smart Wireless THUM adaptors to enable access to HART® diagnostic data available from traditional field devices, we were very familiar with wireless technology. As with the THUM adaptors, the wireless acoustic transmitters were very easy to install, which enabled us to start monitoring instantly. Operators also find it easy to work with the new solution, including monitoring alerts and history for the steam trap state.

Already we have seen the benefit of installing these devices. A leak from a high pressure super heater steam trap was immediately identified that would have cost £1400 for every 24 hours of operation. This includes lost steam when the plant was running, lost pressure when in hot standby mode and also lost nitrogen when the plant is in cold standby mode.

The devices have been installed for only a few months, but they have proven very robust. We had

a leak recently that exposed one of the devices to high temperature steam and it still performed correctly. Any concerns we might have had about the ability of the devices to withstand the tough

environment were quickly forgotten. This is the level of reliability we are looking for from devices

deployed at the plant.

We have now installed fifty devices to target problematic areas by monitoring the condition of steam traps or valves when the plant is running. There is no need to install an acoustic transmitter on every single steam trap. The devices are so portable they can be moved around the plant to where they are needed. Installation requires very little technical expertise, such that a night shift operator can simply move a device from one location to another where a potential problem may exist, without waiting for the maintenance team.

Unexpected Releases from Vent Valves

Having deployed the acoustic transmitters to identify failing steam traps, we instantly saw the potential of these devices and have installed 15 more to monitor other problematic areas of the plant.

Steam loss isn’t restricted to steam traps; for example, vent valves can stick during start-up. We normally have an operator viewing them during this time, but by installing wireless acoustic transmitters we can now monitor these devices from the control room instead, removing the need for field observations. In addition, some of our pressure relief valves don’t seat correctly, which causes steam loss. These valves are located 25m high on top of the boilers and are difficult to check visually by operators; in the past we have tried using CCTV cameras. Manual monitoring for releases occurs periodically, but does not indicate when or why a release occurred, increasing the chances of a safety, regulatory, or environmental incident. Wireless devices enable us to be very precise, alerting operators when the relief valves have opened.

The time-stamped alerts can also be compared against process conditions or environmental reporting to help identify the root cause of a release so that preventive actions can be taken to reduce future emissions. Data is transmitted to a Wireless Gateway and – using OPC – is integrated into our existing Distributed Control System (DCS). Data is fed into the DCS’s Historian, where it is easy for us to trend noise levels and identify changes. Within the DCS, the wireless devices appear exactly the same as any other measurement devices. Not only have we been able to configure alerts based on parameters, we can trend results and make informed decisions based on this data. In both these applications we have gained a greater visibility to our process through improved monitoring.

Boiler Tube Leaks

A further problem is leaking boiler tubes. Tube leaks caused by corrosion and subsequent cracks reduce the performance of the boiler, affecting the overall output and profitability of the plant. Boiler tube leaks waste large amounts of demineralised water – which we have to produce on site at a cost – and also chemicals used in the feedwater to help prevent corrosion.

 Leaks can usually be identified by the amount of steam entering the drains and the quantity of additional water being put back into the closed loop system. Early detection is critical, allowing leaks to be addressed through a planned process before the problem escalates, leading to severe damage to surrounding tubes and tube banks. To repair a boiler tube can take around 6 hours, so we need to carefully schedule the repairs. Forced downtime could cost the company tens of thousands of pounds depending on the unit price of power at the time. Condition based monitoring allows for repairs to be undertaken during normal off-times helping us to maintain maximum plant availability.

 Leaks create considerable additional noise, so we decided to try the wireless acoustic transmitters in this application. To get the system running, we simply had to monitor the ambient noise measurement on the devices and configure parameters to set off an alert. We are also able to trend noise levels to identify gradual changes. This enables us to spot problems early, reducing waste and helping to improve maintenance procedures.

 As with the steam traps and valve applications, installing traditional wired monitoring devices on the boilers would also have required extensive new cable infrastructure as none was available. Our positive experience with wireless technology meant that we had no hesitation in selecting another wireless solution for this task.

 A Foundation for Future Benefits

 With three wireless networks now in place, additional devices can be added anywhere in the plant at much lower cost than if they had to be wired in individually, and without having to worry about the existing cable infrastructure or the need to add to it.

 As part of this strategy, we intend to apply additional acoustic transmitters to monitor the venturi eductor that is installed in the water treatment plant. This device can become blocked by crystallisation of the chemical and thus lead to a failure of the batch process. Should the ejector become blocked, the flow rate reduces or stops and there is a reduction in noise levels. The acoustic transmitter can be set to alert operators should noise levels drop, indicating a blockage. This immediately allows the process to be aborted early, saving vital raw materials and reducing potential effluent discharge.

Improving process performance is all about understanding what is happening and being able to respond quickly. By having the opportunity to introduce additional measurements, we have all the information we need.

WirelessHART is based on self-organizing MESH network.