Using the IIoT to Detect Low Probability/High Consequence Events

By Bob Karschnia, Vice President and General Manager of Wireless, Emerson Process Management

Using the IIoT to Detect Low Probability/High Consequence Events

In the wake of major incidents like the Deepwater Horizon Gulf of Mexico oil spill, large corporations in the oil and gas and other industrial arenas are starting to rethink their approach to safety and accident prevention.

Past practices centered around detection of areas with a high probability of failure. These areas were addressed first, and consequently many companies have done well in terms of mitigating possible incidents from these high frequency failures.

But as Deepwater Horizon and other similar incidents have shown, it’s now time to look at areas with a very low Pump-Motor_Refinery_Day-RGB-med-resprobability of failure, but with extremely high negative consequences if something does go wrong.

By adding additional monitoring parameters such as pressure, temperature, level, flow, vibration, acoustics, etc., it is now easy to see how close a plant or facility is to its target Integrity Operating Window (IOW). Additionally, these measurements can also form Independent Protection Layers (IPL) to increase confidence in the integrity of the measurement system already in place.

Although it’s important to address these areas quickly, there are budget and operational constraints causing companies to look for new solutions. The traditional approach is to install instruments to monitor these process parameters, and connect these instruments to control and monitoring systems via hardwiring. While this approach works well, it can be quite expensive, particularly when retrofitting existing facilities.

Installing the instrument is only a part of the task, with another major cost being providing power to the instrument, and especially signal wiring from the instrument to the control and monitoring system. While the power wiring can often be tapped from a nearby source, the signal wiring needs to run all the way from the instrument to a control and monitoring system, at distances ranging from hundreds to thousands of feet. And at the control and monitoring system, new input points need to be added, which can also be very expensive.

An alternative solution harnesses wireless technologies in an Industrial Internet of Things (IIoT) approach. Wireless sensors are installed at the points of measurement, with each having an integral power module with a life span greater than 10 years.In some cases, these wireless sensors can use energy harvesting devices which scavenge power from the heat differentials often found in industrial facilities, extending service life indefinitely.

No signal wiring back to the control and monitoring system is needed, as each wireless sensor is connected to a gateway via a self-forming and self-healing wireless mesh network. The gateway is then connected to the control and monitoring system via a digital data link such as industrial Ethernet, obviating the need for added input points.

Each wireless sensor also acts as a signal repeater, creating a robust network capable of spanning large areas within a plant or facility. To traverse longer distances, say from an oil production platform to a centralized control and monitoring system, various technologies can be employed such as cellular, RF or satellite communications. Companies can use these communication technologies to create an industrial Intranet of Things, and of course the Internet can also be used as a communication medium to create an Industrial Internet of Things.

To show how this works in practice, consider the monitoring of pumps in industrial plants and facilities (Figure 2). When pumps fail, the consequences can range from minor disruptions to major incidents including fires.

Some of the pump seal systems at the Flint Hills Pine Bend refinery in Rosemount, MN were leaking, which led to the formation of vapor clouds. Heat from motors powering the pumps ignited the clouds in a few cases and caused fires. Although none of these fires had caused a major incident yet, the potential for such an event was too high for comfort.

The plant wanted to address this issue as quickly as possible, and they needed to do so in a cost efficient manner. They implemented a wireless pump health monitoring system to provide automatic and continuous monitoring of 158 pumps, and in the process found this IIoT solution was one-tenth the cost of the previous practice of making periodic manual rounds to check on pump condition. The wireless IIoT solution has been in place for five years, with no pump fires during this time.

Pump monitoring systems such as the one at Flint Hills are built around wireless pressure, temperature and vibration sensors which send data to specialized information processing systems in the cloud or in a central control room. These systems detect conditions leading to problems, and then alert personnel with actionable information so they can schedule preventive maintenance to prevent problems before they occur.

Preventing low frequency but high consequence incidents is a priority for many industrial firms. IIoT implementations using wireless sensors allow these issues to be addressed quickly and in a cost-effective manner.