Improving Plant Performance with Technological Innovation

From software to boiler cleaning, evolving technologies and associated expertise are enabling new and aging plants to improve reliability and optimize plant operations.
By Anna Simet | April 20, 2021

Avast majority of biomass and waste-to-energy plants in service today—particularly in the U.S.—were built many decades ago, largely from the 1970s to the 1990s. Though many of them still have plenty of operating years left, some of the technologies deployed at these facilities need upgrading or replacement. For plants already equipped with modern control systems, hardware and software, O&M advancements are making troubleshooting, maintenance and repair work more efficient and effective, extending equipment lifespan and maximizing system output.

From Symptom to Diagnosis
When it comes to optimal biomass plant performance and combustion, RJM International takes a holistic approach, according to Larry Berg, vice president of engineering. “This means we take the entire plant performance into account, because oftentimes, our experience has been that problems or symptoms in one part of the plant can seem unrelated to the actual problem being seen,” he says. “We always try to understand each plant and its unique operating characteristics—you need to do that in order to determine the root cause of any real problem.”

Berg refers to RJM as a small company of experts on all aspects of plant performance.  “We diagnose problems, design solutions, fabricate hardware, install, commission and optimize,” he says. We all get our hands dirty and go out to make things work. Many large companies have lots of us, but rarely di you get to interact with them.”

When a plant experiencing a problem engages RJM, the first step is a combustion audit, Berg says, during which an on-site evaluation is performed to diagnose the problem. “We may not tell them what the problem is then, but we come up with a preliminary hypothesis of what the root cause is,” he says. “We use CFD (computational fluid dynamics) analysis to verify our diagnosis, and then develop process design to solve the problem, including hardware design, fabrication, installation and optimization in a given plant.”  
As for the challenges of biomass and municipal solid waste (MSW) fuels—including woody biomass, pellets, raw MSW and pelletized MSW—Berg says RJM has seen it all, including poor design, operator error, material handling problems, high emissions, poor reliability, inconsistent boiler output and ash deposition and sliding. “One project had massive boulders inside the combustion chamber,” he says.

Berg emphasizes that any given issue can cause another issue downstream, so the root issue isn’t easily understood or obvious, even though the symptom is viewed as the problem. “For instance, you can have an unreliable fuel source—grate nonuniformity—and at the end of the day, it causes CO, so a plant calls us to say it has CO problems.”
There are many more examples of the root cause not being obvious, Berg says. For example, the root cause of lower furnace slag may be large particles recycling to the bed. The root cause of reduced generation might be poor fuel distribution. The root cause of high selective noncatalytic reduction costs might be an inefficient over-fired air system. The root cause of economizer fouling might be similar—but different—fuel. The root cause of boiler tube corrosion might be that the storage location is near salt.

“It’s important to gain a thorough understanding of the operations of the plant, because the problem doesn’t necessarily tell you what the root cause is—finding it out takes a little more study, evaluation, diagnostic and time on-site,” Berg says.

A specific example Berg gives is an instance in which RJM was tasked with solving performance-related issues at Energy Works Hull, a U.K. biomass power plant. One of the issues the plant was experiencing was difficulty meeting generation capacity. RJM recommended a combustion audit and plant review, and subsequently determined that inconsistent fuel flow and limited throughput was causing the issues. “In other words, the feeders weren’t working properly,” Berg says. “We proposed feeder modifications—there was a choke point we were able to relieve—and we also proposed a unique variable pitch auger that allowed the feed to go more uniformly across the bed. The redesign increased throughput by 40 percent and improved stability, restoring the loss generation capacity.”

 As biomass fuel has some characteristics unique to traditional fossil fuels, thermal imaging advances have allowed operators to gain a clear picture of what’s happening in the boiler—for instance, Ametek Land’s mid-infrared borescope imager.

Seeing Through the Smoke
Thermal imaging is an extremely useful tool for gaining insight into the combustion process in a biomass boiler, whether it be a sloping grate, fluidized bed or pulverized fuel, according to Derek Stuart, global product manager at Ametek Land. “Biomass fuels have some specific problems that don’t apply to more traditional materials like coal and oil,” he says. “They’re often fibrous and sticky, which makes them difficult to feed to the furnace. They tend to have low ash fusion temperature—the temperature at which fly ash softens and becomes sticky—making it adhere to the heat transfer surfaces in the boiler.”

Some biomass fuels such as straw contain large amounts of chlorine, potassium and sulfur, which can lead to deposits on the heat transfer surface causing efficiency reductions, and often, severe corrosion, Stuart says.

Thermal imaging allows a view into the boiler to observe any of these potential issues while the asset is in service. Simply put, it works by detecting the infrared radiation emitted by a warm or hot object, Stuart says. “The intensity of the infrared radiation correlates with the temperature of the object,” he explains. “Conventional thermography uses the infrared image to show which parts are cool and which are hot. Radiometric thermal imaging takes things a step further, and allows the sensor to make accurate measurements of the temperature at each point. This can be especially valuable if there are limits to the allowable temperatures of parts of the object being imaged—for example, if the steam tubes in a boiler can be damaged by getting too hot.” 

Nonthermal imaging—i.e., using visible images—is typically dominated by flames, making it difficult to see what’s going on through the smoke and dust. The IR image, however, uses special filtering to reduce those observations. Ametek’s MWIR-b Imager has a resolution of 640 x 480 pixels and is available in two temperature-measuring ranges, with one covering from 500 to 1,800 degrees Fahrenheit (F) and the other from 932 to 3,272 F. Lengths are available in one, two and three feet, with automatic retraction functions if purge air or cooling water are lost.

Stuart points to a real world example of thermal imaging solving a biomass plant boiler issue: a plant combusting olive pulp, the leftover material after oil is pressed out. “It looks a lot like sawdust and has a sticky texture,” he says. “The boiler is 15 MW and has two grates with two feed chutes. Fuel is fed into the boiler every 45 seconds with the feed alternating through the two chutes.”

 The boiler operator was having problems feeding fuel evenly, experiencing blockages. With thermal imaging, operators were able to see inside of the boiler and observe that the fuel was hotter on the left side of the boiler and cooler on the right side, Stuart says. “One of the key features of any thermal imaging system is the image processing software that allows information extraction. We can emphasize different aspects of the data and choose appropriate areas of interest, using isotherms to highlight areas of hot and cold temperatures.”

The imager allowed a view of a large portion of the boiler interior with little interference from smoke and flames, including the fuel flow in the grate. “We were able to measure temperatures in different parts of the furnace and highlight different areas of hot and cool temperature,” Stuart adds. “And all of this done with minimal interference to the boiler’s operation.”

As for boiler cleaning, KEPS SPG Inc.’s shock pulse generators are an ideal nondestructive option for boilers experiencing buildup and needing a simple solution that can be performed while the plant is online. Mitchell Pezzi, president, says there are 700 units installed worldwide, with about 12 percent at biomass plants, and 70 percent at waste-to-energy plants in Europe and the Far East.

Shock Pulse Generators for Effective Boiler Cleaning
Plant operators deal with many things to achieve their ultimate goal of optimizing combustion and boiler efficiency, Pezzi says. “There are some hindrances that can prevent that. In the boiler, slag buildup can be a very detrimental to boiler efficiency. When it comes to clinkers and slag, the concern is that clinker will build up and fall, damaging the tubes at the bottom or the grates. Poor thermal transfer in general is also an issue. We want to keep the boiler tubes as clean as possible, because as we coat those tubes, the thermal transfer is dramatically reduced.”

Sootblower erosion—caused by mechanical removal of material over time—is another potential issue that causes tube leaks. Pezzi says that shock pulse generator technology can effectively solve these issues without causing any potential damage to assets. These systems work by sending automatic, directed shocks or pulse waves derived from the controlled combustion of oxygen and methane, out of a valve and Venturi nozzle into the boiler. “In essence, what we’re doing is combusting oxygen and methane under pressure to create a high-pressure wave that cleans the boiler tubes,” Pezzi explains. “This pressure wave—the high peak is above atmospheric pressure—is followed by negative pressure back toward the source. We are trying to break that boundary area between the ash and the boiler tube. We’re not blowing it off or pushing it off—we want the pulse wave to fracture that ash. The wave travels in one direction and fractures it, and then comes back in the other direction and fractures it the secondary way. The key to this is that as we’re doing this, we’re moving the air to create a turbulent zone to bounce from tube to tube or wall to wall. The added part is that we get 360-degree cleaning.”
The shock pulse generator is mounted to the outside of the boiler wall and is suspended by a hanger. “It’s a very straightforward, compact system,” Pezzi adds. “We wanted to ensure effective boiler cleaning, and not only do these devices clean deep and penetrate further than traditional methods, a single unit cleans a much larger area and there is no collateral damage—no air, steam or water, no nozzle, jet tube or tube damage.”

As for inside of the plant, many facilities built decades ago have or will soon need control system overhauls for a number of reasons, and there are a couple of ways to approach it, according to Shawn Coughlan, vice president of Applied Control Engineering.

Retrofits for Aging Control Systems
Drivers for control system retrofits are numerous and include hardware and software obsolescence, technical issues and improved operational characteristics, says Coughlan. “Many control systems in biomass plants were installed in the 1970s, ’80s and ’90s and have reached the end of their lives. The hardware cannot be repaired, principally because the components on the boards are no longer available—the resistors, capacitors, etcetera—cannot be purchased. The vendors are unable to manufacture or repair older control systems, and a lot of owners have resorted to buying spare parts off eBay.”

As for technological drivers, processing power has exploded over the years, as well as Ethernet, or connectivity in all components. The control systems built back then had dedicated, custom-built networks and data highway communications, Coughlan says, but today, all of that has been replaced by a standard Ethernet cable, enabling control systems from different platforms to connect.”

Wireless technology has allowed operators to gather much more data from tanks, vessels and other equipment, without the need for extensive wiring. Cellphone processing power has increased immensely, allowing operators to connect to the control system and control it on their phones, Coughlan says.

Security concerns have also changed over the past 20 years, with virus and intrusion protection necessitating new hardware and software platforms. “Older systems can’t be patched, as their operating systems are no longer being maintained,” Coughlan says. “There’s a notion that systems not connected to the internet—i.e., air-gapped—are safe from intrusions, but that’s not the case. At many of these sites, the technicians, engineers and operators bring laptops and connect to the system or plug into USB  ports.”
There is also an increased focus on process safety, and analysis of machine learning has allowed operators to predict equipment failure, rather than wait for catastrophic failure, Coughlan says. Finally, he adds, data integrity in the older systems isn’t well maintained, and new systems are greatly improved. “The traceability, storage and backing up of that data isn’t up to today’s standards. The need for data analytics and dashboarding, providing KPIs (key performance indicators) to the operator, supervisor and plant management—those needs are driving changes in the operating platform, requiring more data to be brought in. Many of the old systems can’t handle that influx of data.”

Coughlan outlines three technical approaches to DCS retrofits—console replacement, logic solver replacement, and I/O module replacement.

Console replacement is the most frequently performed approach, principally driven by aging computers, the spare parts of which are extremely tough to come by for many systems, Coughlan says. This is either because the operating systems are not being updated or that the manufacturer has discontinued them and no longer provides support. “Much of the support for these systems comes from system integrators or retired personnel,” he says. “When a console replacement is executed, it may be a same-vendor upgrade with a straightforward migration path,” he says. “If you switch to a different vendor, you might have to convert graphics from vendor A to B. Fortunately, the console replacement cutover is very simple and can be done with no downtime.”

Logic solver replacement can be done with or without replacing the I/O (input/output) modules. “As with the console replacement, the same or different vendor can be used. The same vendor can usually import the configuration from old system into the new, and will probably get about 90% of the logic without any changes. With a different vendor, you may still be able to import or convert the data, or you may need to reprogram and also replace the I/O.

Coughlan notes that when replacing the logic solver, if the old system configuration is imported, it will bring along with it the old code and all its associated problems. “If you reprogram it, you’ll need a good definition of what’s existing and you may need to reverse engineer, which takes a lot of time,” he says. “You may not have anybody at plant who knows how to do it, so the systems integrator might need to spend a lot of time to decode it.”

While reprogramming takes time, plants achieve the benefits of cleaning up their logic, and a fully documented system. “You’ll fix those nagging logic issues, and you’ll get to use the best languages available today for your programming needs,” Coughlan says.

I/O module replacement may be the most time-consuming part of the installation, according to Coughlan, due to rewiring requirements “Sometimes, we can provide you with an adaptive cable so you don’t have to rewire every point, but you’ll need to match I/O types and I/O counts to make this a seamless transfer. Any time you do I/O changes, you need to do loop checks. If you touch every wire, you’ll need to a loop check on every point. If you use adaptable cable, however, you can get away with testing one or two points on a cable.”

Control system retrofits or replacements can be done using one of two approaches—phased or all at once, which Coughlan refers to as “rip and replace.” Phased implementation has the benefit of spreading it out over time, he says. “The consult, logic solver and then I/O replacement enables cost spread out over several years, and allows you to work within the shutdowns you have.”

Plants can expect a one- to five-year implementation plan for control system replacement processes, Coughlan adds. “There are a lot of technical challenges to both the all-at-once or rip-and-replace approach, so you will need to spend the time, and dedicate the manpower of O&M personnel along with the engineering staff, to achieve the goal.”
Author: Anna Simet
Editor, Biomass Magazine
[email protected]