Wireless mobile alarm implementation
System improves process efficiency and environmental compliance in mineral processing operation
By David Milne
Excellence in integrated mobile alarm management requires attention to control, asset management, productivity, and regulatory issues. All of these factors came into play recently when Invensys Operations Management implemented a wireless infrastructure for a large mineral processing operation. Mineral processing is an extremely challenging industry because profitability correlates closely to demand for the finished goods created, as well as to mineral prices, which can fluctuate wildly year to year based on economic conditions. Therefore, keeping operating costs low without jeopardizing environmental compliance requires real-time attention to operating efficiency. In a large mineral or mining complex, where operations can be dispersed over many miles, effective alarm handling can be very strategic and help achieve operational excellence.
One large U.S.-based molybdenum processing company approached Invensys to implement a system that would allow operators to receive alarms remotely, enabling them to respond quickly to environmental contraindications, while simultaneously monitoring and adjusting process set points. If differing ore compositions required even minor changes, the operators would have to stop what they were doing and return to the control room to make the necessary adjustments. Such intervention had to be balanced against the emissions from the process. For example, a molybdenum roaster building that is nine stories tall would likely have several hearths per roaster, and getting down to the control room and then back to the task at hand can be very time consuming. Delayed reaction to air quality-related alarms, however, can result in substantial penalties, so speed of response translates into dollars very quickly.
The processing yard was equipped with a wireless network that connects the operators to the existing process control system via handheld computers. This enabled them to bring the control board, alarms, and adjustments with them while they worked on other tasks. The solution consisted of a secure network of wireless access points coupled to the process control system via handheld computers using encrypted wireless links. The wireless access points connected via a secure mesh topology to a gateway, which was hardwired to the process control network. The gateway access point to the plant control network was the only wired connection. Everything else connected to the gateway via encrypted wireless backhaul links.
Wonderware SCADAlarm alarm software was installed on the existing process control system. The software directs system alarms to the handheld computers and system console. Operators receive the alarm as either an audible or vibratory signal, and then connect to the system from their mobile units. The control system graphics that the operators see on the handheld computer match the process control system graphics exactly. Instead of the keyboard and mouse, however, operators interact with the process via a stylus and touchpad. The roaster operator, leach operator, and site supervisor each have these mobile systems.
Adding the wireless infrastructure and handheld computers liberates the control room operator to investigate changes in the field, while remaining in touch with the operating system. Previously, leach and roaster operations required a dedicated operator for each console. This individual would acknowledge any alarms, initiate necessary corrective actions, and notify additional operations staff appropriately. Now, one person can work both consoles, and appropriate staff receives notification simultaneously.
In addition to operators and supervisors, plant electricians are also using the handheld systems to determine the status of the equipment they are troubleshooting or installing, forgoing the need for someone else to sit at the main console to monitor the effect of any adjustments they might have to make.
Building the mobile infrastructure
No matter the industry, wireless implementation issues are the same. The typical infrastructure for mobile alarms is usually 802.11a/b/g/n, where the suffixes relate to frequency, throughput, and standards of the wireless infrastructure. This must be implemented in the context of competing technologies and, in some cases, competing purposes, such as voice transmission.
For example, mobile alarm systems need a high-bandwidth, low-latency solution to promote a good user experience and encourage adoption. If a voice requirement is added, then users have to make assumptions on how much bandwidth per access point must be reserved to maintain voice quality without impacting performance. One has to estimate the maximum number of voice users per access point, which might drive a greater-than-anticipated access point density to reserve the requisite bandwidth. And this is important to determine whether to plan for day-to-day operation with phone use acceptable outside the battery limits, or for turnaround use by staff and vendors throughout the facility. Both uses require multi-megabit service rates to achieve expected performance. Different applications have different network traffic characteristics and bandwidth requirements. Voice, for example, is a stream of data throughout the call that, if interrupted, results in gaps in the conversation. As with any network, knowing the characteristics of your network traffic is essential to completing the assessment.
Also potentially competing for bandwidth would be communications technologies that use a different sort of radio modulation across the same spectrum as voice, decision support, and alarming software. One well-known manufacturer of wireless instrumentation, for example, offers a radio combination that uses two sets of frequencies: one to deliver the backhaul and the other to deliver the edge communications with the instruments. The edge connections use radio frequency modulation that is extremely tolerant of interference because the emphasis is on providing the data at speeds less than 115 kilobits. The modulation technique allows for a retransmission to prevent a timeout should the operator encounter interference on the radio frequency. The effect of that would be the lowering of the throughput rate to ensure successful data transmission, which would in turn impact the planned criteria for voice and data.
A lower throughput rate without errors is generally preferable to a higher throughput rate with errors because the higher rate may require many retransmissions. But both compromise the end-user experience because there is a processing delay or degradation of voice quality, which could vary from being barely perceptible to being unusable. It is easy to see how the complexity of a wireless implementation can snowball when you need the wireless system to run alongside an existing corporate wireless infrastructure within the same operating environment or when a neighboring company has also implemented wireless technology on the edge of your plant’s operational area.
Additional sources of interference
Factors beyond spectrum availability may also raise interference issues. For example, metal is a good absorbent and reflector of radio signals, which, when bouncing off metal structures such as piping, process vessels, and storage racks, may result in spurious signals that could reach the receiver at the same time as the true signal, thus creating coverage canyons. To minimize the impact of such “multipath interference,” access points must be situated so they can find the gateway while providing the appropriate area coverage for the handheld computers. Receivers must also be programmed with algorithms to reassemble the data. The more multipath interference the radio experiences, the lower the data rate. Liquids and construction materials generally absorb the signal in varying degrees depending on the material. When the signal is absorbed, the result is a “fade,” e.g., the signal works here, but disappears when standing behind a bulk storage tank.
Storage racks in a warehouse can create additional complications, depending on the purpose of the warehouse. The stored goods themselves can absorb or reflect signals. Network designers need to know, for example, whether the warehouse is storing canned peas, bottled water or cornflakes, and will it be the same mix in the same location tomorrow?
Any proposed wireless solution should be carefully thought out in regard to how the existing infrastructure is going to be extended wirelessly. An effective wireless infrastructure will incorporate security during the design phase, after the network requirements have been established, and not as an afterthought once the network has been put into place. This encompasses more than selecting which encryption protocol will be utilized to secure the wireless traffic. Thought needs to be given to authentication and authorization and adds to the need for a comprehensive design.
An alternative to building a mobile wireless infrastructure is to use alarm notification methods that utilize other transports, such as pagers, short message service (SMS), e-mail, voice, or even the plant’s paging system. Standard alarm software, such as Wonderware SCADAlarm, can be used to access an existing alarm management infrastructure with functionality enabling the users to configure the system to send e-mail messages, SMS text messages, or phone calls associated with the alarm to cellular telephones or pagers. SMS, e-mail, and voice are convenient methods because they can all utilize cell phones in addition to corporate e-mail accounts and telephone systems. Voice functions can broadcast alarms over the plant intercom using a connection to the plant intercom system. They can also select alarm tags and communications paths, set scheduling criteria, designate active versus inactive operators and target stakeholders, such as process owners, individually or in groups.
Beyond alarm management
Once implemented, a mobile wireless infrastructure can be enhanced in many ways. Mobile decision support systems can bring additional procedural guidance and decision criteria to an operator working on remote equipment, and they can be integrated with just about any line-of-business application for the operating system. Just as an alarm can be configured to alert operators to regulatory breaches, they can also signal real-time supply shortages, cost drains, and even profitability impact, helping to drive real-time improvements in operational efficiency. However, one size network or one infrastructure does not fit all, so careful preplanning is essential.
Predictive modeling, fueled by a complete audit of the intended uses of the network, is the best way to avoid these and many other issues in industrial wireless implementation. The model should result in a graphical mapping of the coverage areas by signal strength and should show any probable dead spots associated with the architecture. The methodology should include modeling of the area under consideration, an RF spectrum sweep to locate sources of interference (existing wireless, electric motors, equipment in use at the plant next door), and a validation of the generated model. Such validation would consist of placing access points near the locations proposed by the model and verifying that the expected signal strength received by the equipment is close to the model’s prediction.
While this will require a little more planning and effort than setting up that wireless router that you have pulled out of your basement, the performance, business, safety, and regulatory benefits achieved will pay off the effort many times over.
ABOUT THE AUTHOR
David Milne is an applications engineer in Invensys Operations Management’s critical infrastructure and security practice. He joined the company in 2007 and has more than 15 years of platform and network experience.