Another interesting application of wireless sensors and mesh networks
posted on
May 19, 2009 10:52PM
We make wireless work.
FYI...
Midstream
|
May, 2009
|
By Graham Chandler
As morning coffee drinkers know, getting too wired can reduce productivity. The same is true for sensor systems in drilling rigs, gas plants, and many other processes. Running wire is always an expense, and sometimes very difficult to achieve. The solution? Take a look at WirelessHART (Highway Addressable Remote Transducers), an industry communication standard. This technology is to wired sensors what cellphones and BlackBerries are to land-line telephones-if anything, WirelessHART is even more revolutionary in the petroleum sector and far beyond.
A pioneering player in this emerging field is Dust Networks, co-founded in 2002 by its chief technology officer Kris Pister. His team came up with miniature transmitter and sensor assemblies that sit inside wireless monitors. Pister refers to these assemblies, as small as a grain of rice, as motes, or as smart dust. He envisions a world of inter-communicating mini-sensors scattered like dust, gathering information economically and reliably. "In the future, every square inch of every city will be alive with intelligence," Pister stated in his presentation to the International Society of Automation at its ISA 2008 conference in Houston last October.
WirelessHart has attracted a global bevy of industrial automation manufacturers, including Honeywell, ABB Group, Yokogawa, and Invensys. Emerson Process Management, which supplies Calgary-based Spartan Controls, is an automation heavyweight (2008 sales: US$6.6 billion) that unveiled wireless technology in 2006. The wireless aspect is an important new innovation that delivers key economic benefits, says Spartan technical sales specialist Neal Tetreault. "For some temperature measurements in plants, it's sometimes not feasible to run wires all the way up to the stack," he explains. Now an operator can install a wirelessHART device up that same stack to achieve the same result, possibly with a cost saving as high as 90 per cent.
The Emerson/Spartan Smart Wireless solution consists of wireless gateways and field instruments, also called devices. (Those are "blue cans" in the photos.) The field device has two functions: 1) To receive and condition data received from sensors that measure temperatures, pressures, flow rates, vibration, and other process parameters; 2) To relay the measurement data to the gateway. For years, the heavily "ruggedized" field devices were wired. Now, up to 100 of the devices can be linked wirelessly to a gateway, which passes on the process information to the control system of the plant, drilling rig, or other industrial facility.
Smart Wireless has another extraordinary feature-once installed, the technology deploys and optimizes itself. "The network self-organizes, you don't have to do anything. Just start 'er up and away you go," says Tetreault. The mesh network optimizes itself at need, constantly staying in contact with the gateway and implementing reconfigurations in response to problems. Some call this capability self-healing.
Even while the mesh network adjusts, the gateway passes all sensor data to the control infrastructure using one or more options from various communications protocols. Multiple networks may be bridged using Wi-Fi, "just like at an airport, for instance," Tetreault says. "We teamed up with Cisco to create plant networks based on IEEE 802.11 which is Wi-Fi. It's a highly ruggedized version of the technology you're using at home in your routers." In addition to bridging the WirelessHART networks, the wireless plant networks empower mobile workers for greater productivity, bridge isolated control systems, improve safety mustering, track personnel and assets within expansive plant sites, and accommodate non-traditional signals such as video. The self-organizing mesh capability comes to the rescue if a wall or a pipe or even a building blocks "point-to-point" radio transmission. (A common point-to-point device is a cellphone, which will lose a call in blind spots where it cannot directly "see" a transmission tower.) If a WirelessHART field device cannot address the gateway directly, it will talk to another wireless instrument which in turn relays that data on to the gateway.
"So now if you have obstructions in the way-every plant has them-the more devices you install, the stronger your network," Tetrault says. "If you put a building in front of it, it just goes to the next device to the gateway." If one sensor goes down, the signal simply finds an alternate path. As conditions change or fresh obstacles are encountered in a plant, such as temporary scaffolding, new equipment, or a parked truck, WirelessHart networks simply reorganize to get their signals through.
It's also great for drilling rigs, says Tetreault. "For all your temperatures and pressures on the mud tanks, your BOPs [blowout preventers], and it's ideal because now you don't have the wires." When rigging up and rigging down, "you don't have the issue of having to hook up and disconnect all the wires," he points out.
The networks are finding extensive use in highly automated and multi-process plants like refineries and upgraders. In these industrial settings, dozens or hundreds of sensors are needed. The technology becomes almost self-expanding: savings on the wireless installations constantly allow a company to afford to deploy more sensors in places they previously could not cost-effectively deploy wired sensors, or could not practically install them due to location. It all translates into improved productivity, safety, and reduced plant downtimes.
Talisman Energy's Teepee Creek sour gas plant near Grande Prairie used WirelessHART to address safety issues within its process. "We didn't have the ability to see pressure inside the tanks, they were a long way from all the other process piping." says Tyler Schmidt, Talisman's automation lead in the Grande Prairie region. The company needed a way to get pressure indication and trending for the tanks, but the cost of cables and wiring them in traditional ways would have been too high. "The wireless transmitters were an easy fix," the technical specialist says. "That was the first sensor, we put that one in last fall. We have six running there now."
The learning curve continues. Schmidt says the plant uses the wireless sensors on 'indication-only' applications but not yet for process control. "We have to go some way to prove them for a sour gas plant," he says. "I can see it being connected to process control eventually." For a wellsite, the Talisman technician thinks control applications would be ideal. "You can easily back things up there, just flow control or simple control where you have backups in place. We may not use them in this facility yet-we have strategic control, wired. But on a wellsite, it wouldn't be a big deal. It can shut itself down if anything goes wrong."
Talisman's Teepee Creek plant currently uses the HART network to measure temperatures, pressure, differential pressure, tank levels, and one for valve position indication. Schmidt sees that expanding soon. "We'll be using more, probably some on valve indication applications so when they close a valve it will tell the operator it actually closed," he says. "They are proving reliable, so we're confident to go forward and put more in."
Talisman managers appreciate the cost savings. "On a simple wiring application, the transmitter would be just under $2,000 and close to another $2,000 to wire it in. We put in a wireless one for well under $3,000," Schmidt says. Wire itself is cheap, he adds, but the labour for installing it costs a lot. And the Grande Prairie specialist points out that wireless savings will be much more substantial where transmission distances are longer. "On wellsites, it will do really well." For some applications in large plants, wired installations can cost $10,000 or more per sensor.
Savings accrue in power costs, too. Because WirelessHART devices transmit for just milliseconds at a time and each signal has only a few metres to travel, power requirements remain low. "In a temperature transmitter, a battery will last seven to eight years," says Tetreault. Emerson adds that batteries have a theoretical life expectancy of a full decade or more in some applications.
The more sensors, the more efficient the system becomes overall-more pathways for the data being transmitted. Because it's a new technology, customers may buy just one or two as a test project. "We're finding that those customers who buy one or two stretch out and say, 'Hey now, we've got that wireless there, we've always wanted to make this other temperature measurement but never did it because it was too expensive. Now we've got the gateway, let's throw in another transmitter.' Then all of a sudden they're up to 20 or 30 devices or more," Tetrault says.
Process automation companies and battery manufacturers continue their pursuit of smaller, longer-life batteries-a key to increasing measurement and control sample rates in field devices and valves and achieving more extreme miniaturization. Another way to extend battery life is through developing micro-electromechanical transmitter and sensor assemblies that consume less power.
Self-powered smart dust opens many new opportunities. For instance, added into a pipeline or product flow, these mini-sensors could detect leaks through monitoring factors such as volume, viscosity, temperature, and the surrounding environment. Data would be integrated through wireless mesh networking technologies that allow the motes to communicate with each other and synthesize the overall situation.
At the ISA conference, Pister acknowledged that earlier wireless sensor initiatives had often failed, typically due to difficulties with point-to-point. He then reviewed mesh sensor pilot installations in North Sea oil platforms (StatoilHydro), production wellpads (Chevron), and refineries (BP, Shell, and Chevron), as well as other industries. Based on those successes, the Dust Networks CTO asserted that the technology has come of age for applications ranging from patient monitoring to irrigation. His conclusion: "Everything you want to know, you have the ability to know."