The oil and gas industry makes extensive use of heat exchangers, which are at the heart of production of a host of chemicals, petrochemicals and oil products.

Thanks to advancements in remote monitoring technology, the cloud and artificial intelligence (AI), it is possible to improve the approach to monitoring the performance of this important technology. At this juncture, petroleum professionals and other industry stakeholders who are quick to grasp the benefits and applications of the latest approaches stand poised to pull ahead of their competitors.

mCloud Technologies’ AssetCare combines Internet of Things (IoT) sensors, AI and the cloud in managed services to optimize asset performance, minimize energy use and optimize maintenance.

The company recently announced plans to bolster its AssetCare system by acquiring kanepi, a technology company known for its visual analytics system that improves operating efficiencies in asset-intensive industries such as oil and gas. The acquisition allows mCloud to connect tens of thousands of assets and workers with new solutions, and the two companies have already begun integrating their technologies to optimize heat exchangers at oil and gas plants.

Avoiding heat exchanger fouling

Today, there are many kinds of heat exchangers, which come in all different shapes and sizes. Some are as big as a room, while some fit comfortably under the hood of your car. In the oil and gas industry, most heat exchangers are stationary equipment without moving parts. Regardless of the type, fouling–the build-up of scaling, corrosion or other debris–is the bane of heat exchangers.

The way in which heat exchangers foul over time works much like a shower in a house. Hot water is less capable of holding on to dissolved materials than cold water, resulting in calcium or other mineral deposits being left behind. Similarly, heat exchangers used to work with petrochemicals will see a buildup of material over time, even if there is no direct contact between the fluids they’re working with.

This is often a very slow process, taking years for the build-up to become noticeable. In other cases, such as when a facility has changed operation in some way, it can happen much more quickly. For example, if a plant were to switch grades of crude oil to one with higher sulfur, asphalt or an otherwise heavier crude–say from Venezuela or California–it could end up being much harder on the equipment and result in faster fouling.

While this typically doesn’t have any effect on the final product, fouling results in reduced efficiency, which can and end up costing a facility a lot of money. Consider a functioning, new heat exchanger that has 100 units of heat transfer ability. As the unit fouls on one side or another, its effectiveness–the ability to transfer heat–is reduced. Perhaps after a year it drops to 90 units, then to 80 after two years. At some point the exchanger will lack the capacity to perform its job and the facility must work around the problem, perhaps by reducing production or perhaps by shutting down and cleaning the exchanger.

In cases where a particular heat exchanger is the plant constraint and every additional unit of heat transfer allows making more product–and profit–these percentage changes in output from basic fouling can be very expensive, adding up to thousands of dollars a day. Many plants will continue operating under these conditions for years on end, often with many exchangers per site, and the cost starts adding up very quickly.

Although fouling is a serious issue for heat exchangers and, by extension, the plants they serve, they are not always monitored. However, once a plant is sufficiently equipped with sensors positioned to monitor flow, temperature, pressure, density and other factors, mCloud’s technologies can construct a digital twin and compare the mathematical model to what the exchanger is actually doing. This enables the tracking of the heat exchanger’s performance–the amount of heat transfer and fouling–in real time, which is done using three key performance metrics:

Performance. This is current heat transfer divided by heat transfer under initial or clean conditions. There are several ways to calculate the amount of heat transfer, such as measuring the flow rate of one of the fluids and the change in its temperature. For example, a hot water heater that warms 50 gal of water from 68 F to 120 F takes an easily calculatable amount of energy. The best way to do this calculation depends on available sensors.

Once the heat transfer is determined, it can be compared to the design specifications or a first principles process simulator to determine how well the exchanger is working compared to when it was new.

Health index. This index indicates the remaining useful life of a heat exchanger. This estimate is made through a full analysis of the state of heat transfer, flow rate, what the process needs to do, and many other elements to estimate the remaining useful life of the exchanger and more generally, when a heat exchanger will no longer be able to carry out its expected functions.

Utilization. This metric evaluates how much of the heat exchanger’s capacity is currently being used. For example, a plant may be working with an oversized exchanger and only need half its ability to transfer heat for the current operation. Utilization provides an indication of which exchangers might be close to being critical to plant operations, as it can happen that a clean, high-functioning heat exchanger is operating at close to capacity, and is more need of attention that a fouled exchanger with plenty of spare capacity.

Using these three metrics, aided by data analysis via AI, an overall score is then allocated for any given heat exchanger. This score can be updated every few minutes, then monitored against expected values. All of this information is fed into an AssetCare dashboard accessed by operators. mCloud wraps this technology in a managed service, where its Live Operations (LiveOps) staff monitor exchanger performance and receive alerts from the software when problems are detected. LiveOps provides monthly reports summarizing exchanger performance and issues, tracking performance on behalf of customers so that operators don’t need to be constantly vigilant.

Shortly before signing up with mCloud, one of its customers had a heat exchanger unexpectedly foul in the middle of a cold Canadian winter and was unable to produce the needed amount of product. The emergency nature of this maintenance and shutdown–in -35 F–spiked the price of repairs from $60,000 to nearly $120,000. With better information about the state of this exchanger, the operators could have altered their maintenance schedule and avoided the extra cost.

The true game-changer for operators is the ability to do condition-based maintenance instead of running to failure or sticking to a fixed schedule, and to know better how to optimize production and reduce energy costs. Digital transformation has provided new ways to put eyes on heat exchangers. The next leap in optimization is already here, and the plants that incorporate this digital oversight now will save in the long run.