Innovation Honored: AI Control, Advanced Coatings and Leak-Proof Pumps Earn 2025 Vaaler Awards

Sherwin-Williams: Heat-Flex AEB

“Corrosion under insulation is a widespread maintenance and safety issue,” noted one Vaaler Award judge when evaluating Sherwin-Williams’ Heat-Flex AEB insulation coating entry. Another judge also remarked that corrosion under insulation, or CUI, is a significant issue in the chemical industry, adding that Heat-Flex AEB shows promise for eliminating the problem.

Heat-Flex AEB works by building a thick film of insulative coating material onto assets that are required to maintain operating temperatures up to 350°F (177°C), with excursions to 400°F (204°C). This coating retains process heat inside coated assets, allowing them to continue operating in extreme environments. (Figure 1)

Sherwin-Williams touts the technology as an alternative to traditional insulation systems that have the potential to absorb and trap moisture that infiltrates their exterior cladding. Moisture infiltration can reduce the insulating capacity of the insulation and contribute to the acceleration of corrosion under insulation.

The product is a waterborne acrylic that the company formulated along with various engineered particles, which provide the insulation properties, said Neil Wilds, global product director for CUI testing at Sherwin-Williams Protective and Marine. The product can be sprayed onto the equipment, applied by hand or by a hopper gun or texture spray pumps, Wilds said.

Traditional insulation approaches include the use of a mineral-based wall along with jacketing on the outside.

“This is where the problem occurs, and this is where the demand is coming from because to install something as complicated as that on valves and complicated pipe structures, you'd never get the thing fully waterproofed,” Wilds said.

Rain can enter from various sources, including rain, seawater or drift within the plant. This eventually leads to CUI. Through voice-of-the-customer discussions, Sherwin-Williams CUI team determined the problem was prevalent throughout process industries, including the chemical and oil and gas sectors, Wilds said.

“Basically, they needed to get rid of traditional insulation systems and that jacketing, which is holding the water in,” he said. “And this is where we actually started the formulation of this type of material.”

Heat-Flex AEB is Sherwin-Williams’ latest generation of insulation coatings. The company also won the 2023 Vaaler Award for its Heat-Flex line. The first generation of the insulation coating was designed for personal protection or resistance to burns. But customers demanded a product that could eliminate CUI.

Wilds was part of a dedicated team within Sherwin-Williams tasked with dramatically improving the performance of its insulation coatings to maximize protection against CUI.

“The whole project was to try and get away from where we were where we had huge CUI problems and products that didn’t meet the clients’ needs to get to the point where we were meeting everything that the client wants,” Wilds said. “In the end, it’s this CUI problem that’s the issue. If you eliminate the ability of water to get to the steel and you still have the insulation properties equivalent to the mineral wall of a calcium silicate, then you have completely upped the game.”

Wilds added that the company manufactures the product in the northern United States using a trusted formulator that has honed the process of producing the single-component thermal insulative system.

Yokogawa: Factorial Kernel Dynamic Policy Programming

In 2022, ENEOS Materials Corp. announced it had achieved a global first. The company used artificial intelligence to autonomously control its plant in Yokkaichi, Japan, for 35 consecutive days. An AI protocol called Factorial Kernel Dynamic Policy Programming, or FKDPP, was the key technology that enabled the autonomous control. It’s an algorithm jointly developed by Yokogawa Corp. and the Nara Institute of Science and Technology.

Yokogawa describes FKDPP as a technology designed to complement manual and conventional control methods like PID and APC. The algorithm learns optimal control directly from operational data instead of relying on pre-built process models. This makes it highly adaptable to nonlinear, complex and changing process conditions. (Figure 2)

FKDPP can run up to 30 experiments per second.

“The key to FKDPP is how quick it can learn, as one of the [Vaaler] judges already stated,” said Karthik Gopalakrishnan, a technical consultant for digital transformation and smart manufacturing at Yokogawa. “[Another] key of that technology is to make sure there is no overdependency on simulated accuracy.”

Instead, the technology leverages historical data of the processors to optimize control. A reinforcement learning agent, which in this case is the controller, interacts with the process or the equipment that it’s designed to control, Gopalakrishnan explained.

This interaction teaches the AI program how to control the systems through trial and error.

“Predominantly, the whole goal here is to make sure it knows the pattern and it knows how to react based on its behavior of it,” he said.

The algorithm works on a “punishment” and “reward” system, whereby the programmers have coded a boundary into the application that it cannot cross.

“So there are some good guard rails defined for the control parameters,” Gopalakrishnan said.

The system then automatically course corrects itself if it moves beyond those boundaries.

“So that is how we look at it as a punishment,” said Gopalakrishnan.

Yokogawa put the technology into practice at the ENEOS plant’s distillation column, which produces butadiene, a raw material used in synthetic rubber.

Liquid levels in the column and the waste heat and steam require tight controls to separate and extract high-purity butadiene. But the existing advanced process controls and distributed control systems were struggling to adjust two key valves needed to keep liquid levels stable within the column.

“What it really meant was the operator had to closely watch that particular process and constantly change the control valves and the set points for the for that particular controller,” Gopalakrishnan explained.

The AI controller based on FKDPP maintained the fluid in the distillation column at an appropriate level to ensure that no off-spec products were produced. It also operated the valves to heat the distillation column using waste heat as much as possible, resulting in energy savings.

ENEOS has been running the AI-enabled controller for three years and is looking at ways of expanding its use, Gopalakrishnan said.

Flowserve: INNOMAG TB-MAG Dual Drive

Flowserve solved a problem with sealless pumps that made them susceptible to leaking. Chemical plants processing toxic and hazardous fluids, such as hydrofluoric acid or acrylonitrile, often use pumps with a secondary containment system to house liquid in the event of a pump failure.

In the past, canned motor pumps (CMPs) were the only viable option because sealless pumps with dual control systems include parts that wear over time and are susceptible to leaking.

In March, Flowserve introduced the INNOMAG TB-MAG Dual Drive, an alternative to CMPs and other sealless pumps with dual-control systems because the liquid and drive sections of the pump are both sealed airtight, providing a true secondary layer of protection in the pump. The pump is coated internally with a non-metallic liner that features an additional layer of corrosion resistance for chemicals. It’s capable of handling fluids with up to 30% solids in volume concentration. (Figure 3)

While CMPs are also hermetically sealed and have secondary containment, the barriers are created by the motor, not the pump, said Nick Rentzelos, director and general manager at Flowserve.

“With dual-drive, secondary containment is formed entirely by the pump to avoid the inherent safety risks of using an electrical device for liquid containment,” he explained. “So the motor, the junction box, the electrical distribution and control systems are all completely isolated, leaving no place for the hazardous fluid to escape.”

Dual drive also has the advantage of being modular, so users of the technology can add an extra fully redundant layer of protection, said Rentzelos, inventor of the winning technology.

“Our goal with dual drive was not just to provide a pump with secondary containment, but to make it easy, fast and cost-effective for operators to transform their existing chemical process pumps from safe to fail-safe,” he said.

In addition, the dual-drive technology eliminates the need for compressor replacement, shaft alignment, impeller adjustment, emissions testing and oil changes, Rentzelos said. It uses hydrodynamic bearings made from silicon carbide that don’t wear out.

These maintenance considerations are becoming increasingly important as manufacturers struggle with a skills shortage, he said.

“I would say it's the most important,” Rentzelos said. “When we work with our customers, we try to explain the potential ways that containment can be breached and what the ramifications can be, and specifically how products like dual drive are able to provide preventative measures to make sure that things are fully contained, not only when the equipment is working [but] when things go wrong.”

The product design of the pump and motor have an energy-savings upside, as well. The INNOMAG design does not include axial thrust bearings, so the pump operates with less friction and less internal recirculation. It also uses high-strength non-conductive carbon fiber containment shells instead of thin metal cans.

“Metal cans are not only weaker but also less efficient due to the heat and motor drag created by eddy currents and because of the large gap required between the rotor and stator on the motor side to accommodate that can,” Rentzelos explained.

About the Judging Process

An impartial panel judged the Vaaler Award entries. It consisted of the members of Chemical Processing’s Editorial Board and one member of CP’s editorial team. The panel scored all entries on three criteria: technical significance, novelty or uniqueness and breadth of applicability. To qualify, products must have been commercialized in the United States between May 2023 and May 2025.

The judges from this year’s Vaaler Awards include Phil Kaiser,  cyber defense analyst at Dow; Frederick Gregory, process safety and risk manager for the Deer Park and Bayport sites at Lubrizol Corp.; Darren Moroziuk manager of manufacturing support for Pfizer in Kalamazoo, Michigan; Ellen Turner, director of strategic marketing and business development for Andronaco Industries; Julie O'Brien, former corporate sustainability manager at Air Products; and Seán Ottewell, Chemical Processing’s editor-at-large.

The judges selected the winners based on a scoring system in which they rated the entry criteria on a scale of 1-10.

In their comments, the judges provided honest assessments of the technologies and did not always agree on their applicability or effectiveness in industry. However, the judges were overall impressed by the potential each technology holds to enhance both safety and efficiency in chemical processing facilities that handle toxic and corrosive materials.

The next Vaaler program is 2027.