Real-Time Profitability

Whether the business goal is refining the most oil or bottling the perfect batch of beer, the most important function of automation is to help companies achieve optimum performance. But “optimum” is a moving target, and usually dependent on who is defining it. The business variables associated with manufacturing and process plants are changing faster than ever before. In the United States, changes to the power grid, volatile prices for raw materials and fierce global competition mean two weeks is too long to wait to change a process.

Luckily, technology is enabling faster business decision making based on real-time data. Pilot projects are demonstrating value quickly and are being rolled out across locations. And plant managers and controls engineers are on their way to becoming not just efficiency experts, but essential business partners.

For years industry has been focused on real-time performance and efficiency of industrial and manufacturing operations, says Peter Martin, vice president of business value generation for Plano, Texas-based Invensys Operations Management (http://iom.invensys.com). “Controlling the efficiency of an operation directly translated into controlling the profitability of an operation: If you reduce energy consumption, you reduce energy cost. If you increase volume, you increase production value,” he says. “But all of a sudden the business variables in manufacturing have started changing from being highly stable over long periods of time to starting to demonstrate real-time variability. So now the question is how do plant managers and controls engineers deal with a real-time world [beyond the plant floor]?”

The best example of real-time business variability is energy costs, says Martin. It’s all really all changed in the last five to six years, he says, largely because of the opening of power grids throughout the world. With electricity, for example, “ten years ago most of our customers had developed contracts for their energy for as long as a year, and the price of electricity would be a constant over that contract period. Today in the United States, the price our customers are paying for electricity changes every 15 minutes. In the U.K., it’s every 20 minutes. In Spain it’s every five minutes.”

What that means, says Martin, is that “if you reduce your consumption of electricity, your electricity bills might actually go up if you’re consuming electricity during the high-cost periods and not the low-cost periods.” And there is a domino effect on other types of energy, such as natural gas, as well as on some materials, on production values like motor usage and even on carbon emissions.

What many are realizing is that the only way to control such variables is by creating a real-time information environment that enables operations to respond in the same timeframe. And since it’s true that you can’t control what you don’t measure, many real-time data pilot projects must start at the instrumentation and sensor level.

“We’re finding you can start as small as a single process unit, prove the concept works and then roll the strategy out,” says Martin. “You can do it in a very incremental, protected, cash-flow-positive way, then use the cash generated to fund the second piece.”

GlaxoSmithKline plant

At its plant in Cork, Ireland, pharmaceutical manufacturer GlaxoSmithKline (www.gsk.com) has begun a project to better understand water usage throughout the plant. “GlaxoSmithKline is continuously looking to improve plant performance by increasing the number of parameters measured,” says Emmett Martin, site services & automation manager. “Water is a considerable overhead to the plant, so it is important that we monitor flow rates to manage consumption, and to help identify any usage trends.”

The Cork site produces a range of bulk active ingredients for use in the formulation of prescription drugs. The existing water storage facility was too small and had no measurement instrumentation in place. When two new storage tanks were installed along with a new pipework infrastructure, the company also installed ten Smart Wireless flow and pressure transmitters from Austin, Texas-based Emerson Process Management  (www.EmersonProcess.com) .

Using six Rosemount pressure transmitters, two Rosemount flow transmitters and two Rosemount level transmitters, flow data is transmitted wirelessly every 30 seconds, and pressure and level data every 300 seconds to a Smart Wireless Gateway positioned on the control room roof. The gateway is connected using a serial connection to the Emerson DeltaV digital automation system that controls the plant utilities. From there, the flow and pressure measurements are sent to a data historian and made available to plant operators for regular monitoring and reporting.

The new data obtained has enabled GlaxoSmithKline to clearly identify water usage for different areas of the plant, says Emmitt Martin. This provides a far better understanding of the costs, as well as an ability to identify changes that could be made. In addition, the wireless infrastructure makes it easy and cost effective to add additional measurement devices later, he says.

JEA water treatment plant

Cost savings and process improvement are also being found with real-time motor monitoring and management. Running a motor without such data “is like running a 200 mph race car with just a common automobile dashboard’s fuel, speed and temperature gauges,” says John Burns, product manager of control components and system engineering for Alpharetta, Ga.-based Siemens Industry, Inc. (www.usa.siemens.com). “Greater visibility into motor operations and operating conditions enables processes to become ‘situation-aware.’ This means that a process can automatically back off its throughput anytime adverse motor performance might require.”

Burns cites JEA’s choice to install a system to gather real-time motor data for a pumping system controlling the 1,273 lift stations that keep wastewater flowing through 14 treatment centers. JEA is the largest community-owned utility in Florida and the eighth largest in the United States.

According to Darren Hollifield, JEA manager of access and control systems, JEA’s existing pumping control system is outdated and not able to deliver decision making information quickly. Verifying communication between the data concentrators and the remote lift stations “was primitive,” accomplished by checking a counter. Communications are only updated when the Modbus polling device requests data, and the more remote terminal units (RTUs) on a channel, the slower the update times.

Working with JEA team members, Siemens developed a solution in which a Simatic S7-300 programmable logic controller (PLC) provides local logic as well as data collection for a Sinaut ST7 telecontrol system using Profibus for communication. The system enables fully automatic monitoring and control of the lift stations from the master control center, and provides event-driven process data transmission between each lifting station PLC and the control center. The S7-300 is also the Profibus master for two Simocode Pro intelligent motor management systems that provide motor overload protection, local input/output (I/O) to read in the analog well level, and the digital floats’ signals. Simocode Pro provides key diagnostic and operational data that is eventually sent to the control center, including power measurement data that is used to evaluate energy savings projects and control strategies, says Hollifield.

Now, Hollifield says, technicians can see connection failures, disrupted central processing units (CPUs) or a disrupted control center on personal computer (PC) displays that even update automatically after a problem is corrected. Energy savings are realized by optimizing hydraulic control. Pump stations are controlled so that they run in concert and are not pumping against each other. Simocode Pro can measure motor current, voltage, apparent power (kVA), active power (kW), power factor (%) and power consumption (kWh). The system automatically compiles complete diagnostic logs of all communications along with details of the type and time of any failures.

“Simocode Pro provides us with peace-of-mind because those pumps need to be working 24x7,” says controls technician Pat Harwood. At the same time, both he and Hollifield are excited about the potential to optimize their control system’s operations to achieve greater efficiencies, lower costs and save energy.

Rayong Olefins petrochemical plant

Energy efficiency and energy cost curtailment are becoming a high priority in many places, but it’s particularly urgent for heavy energy consumers like oil refining, petrochemicals, paper, cement, and steel production facilities. “Different surveys carried in this area reveal that there is a potential of an average 15 percent (+/- 4 percent) improvements in energy efficiency if ‘best practice technology’ were implemented,” says Selvaraj Sankar, head of the Energy Solutions Research Centre, SGDC, Yokogawa Electric International Pte Ltd, Singapore (www.yokogawa.com). For that reason, these industries have gone beyond point solutions.

Comprehensive industry-specific software tools and templates exist to help companies implement best practices for use of real-time data. Sankar says Yokogawa’s Insight Suite Asset Excellence software includes diagnostics tools such as Fault Monitoring Diagnostician (FMD), which uses advanced statistical analysis methods to help monitor performance at petrochemical companies.

One company with a pilot project is Rayong Olefins Co., Ltd. (ROC), the second largest petrochemical company in Thailand, which produces 1.2 million tons of olefin products annually. InsightSuiteAE was used to collect and analyze process data, and to perform a high-precision simulation of the energy-saving effects using multivariate analysis technology. Energy saving diagnostics were performed against heat exchangers and ethylene cracking furnaces, two of the most energy-consuming pieces of equipment, says Sankar. The resulting savings generated were significant.

As part of its VigilantPlant Service, Yokogawa engineers worked with Rayong personnel to identify normal operating conditions and create reference models using available data. The online measurement was compared against the reference model to understand the deviations in the process and the measurements. “A typical cracker furnace has more than 88 measurements [that are possible]. But traditionally the performance monitoring is limited to ‘efficiency’ calculations and ‘Excess CO2’ analysis,” says Sankar.

Ethylene cracking furnaces have several coil tubes inside through which ethylene raw materials (such as naphtha) flow and are thermally decomposed. Over time, the coke buildup inside these coil tubes lowers the furnaces’ heat transfer efficiency, says Sankar. This increases their energy consumption, necessitating regular removal of the coke buildup. “Without the means to measure the extent of the coke buildup, ROC must use equal amounts of steam to decarbonize all coil tubes, regardless of the extent of the buildup for each. This means that more steam is used than necessary for the coil tubes with only small amounts of buildup,” says Sanker.

InsightSuiteAE was installed to perform multivariate statistical analysis on 220,000 operational points, and then the appropriate amount of steam was made available for each coil tube according its measured coke buildup. For this control, Yokogawa proposed the steam supply optimization algorithm of Exapilot, its operation efficiency improvement package. Application of the solution resulted in substantial savings related to the energy used for decoking, and also prolonged the coking cycle due to improvement in the cleaning efficiency.

Overall, says Sankar, the diagnostics revealed that, for every single unit of heat exchanger and 13 units of ethylene cracking furnaces, Rayong Olefins Co. could expect the following energy savings per year: 807,000 kWh of electricity (and 450 tons of CO2), 1,700 tons of steam (270 tons of CO2) and 300 tons of fuel (800 tons of CO2).

Real-time decompositions

Invensys Operations Management has been working on the concept of real-time accounting information and translating operational savings into real-time business benefits for years. It even developed patented approaches in the early ‘90s. “We were kind of hoping that we could develop something like a software package for the paper industry or a software package for refining that would just connect into instrumentation, we could apply some algorithms and all of a sudden you have all the data you need,” says Invensys’ Peter Martin. “What we learned is, that’s not the case. Not because the software was inflexible, but rather because, if you go into two paper mills owned by same company, even with equipment from the same company, their instrumentation is completely unique. Instrument engineers develop their own instrument approach in every plant.”

So Invensys has gone the services route too, helping companies identify essential real-time information. Martin says Invensys consultants go on site and do four “decompositions”: They analyze and identify the details of a company’s business strategy, accounting, human resources and technology functions. “From that we can identify what instruments are needed, how we convert that instrument data into financial data or into KPI [key performance indicator] data—we sit down and algorithmically or by equation determine that.”

Then, “since the control systems are really nothing more than little real-time algorithm execution devices,” says Martin, “we build a little real-time algorithm that’s designed to do real-time accounting rather than PID control. All of a sudden, you’re exposing real-time data right in the control system, but the data is operational KPI data.” Once you’ve got that data, you can provide it to operators and automatic workflow engines to guide the operators in how to use this information in a more effective way, he adds.

The example he cites is Sasol, a South Africa energy producer, which introduced new technology to help management, operators, and engineers better optimize the energy within the plant, identify the amount of energy needed to meet internal requirements, and minimize its impact on the local electrical grid.

Related Feature - ERP Is Reaching The Shop Floor
To read the feature article relating to this story, go to www.automationworld.com/feature-6427

Providing Sasol operations personnel with a tool that provides feedback as to which boiler produces the least expensive incremental steam enhances performance of each steam station. And, the results have been significant. Producing steam in Steam Stations 1 and 2 resulted in a 6 percent savings in energy and 4 percent savings in electricity costs within the first month (approximately $230,000 US savings in the first month). Energy and electricity savings progressed and improved throughout the second and third months, saving $400,000 US in the first two months from two out of the five targeted plants. The annualized direct benefit of this project was initially expected to be a 2 percent reduction in variable costs associated with energy feedstocks and electricity, but the results have far exceeded these estimates, says Martin.

Helping the craft brewer

Smaller companies also benefit from real-time information. Full Sail Brewing, a craft brewer in Hood River, Ore., needed to leverage a new, automated mash filtration system to gain insight into manufacturing data for greater brew consistency, quality and quantity. As part of the project scope, the company used Portland, Oregon-based Aurora Industrial Automation and its partner, Rockwell Automation (www.rockwell.com), the Milwaukee automation suppler, to install and start-up the new systems. Aurora provides complex industrial control services to the pulp/paper, life science and packaging processing industries.

The brewer had been using a primarily manual process to mix, filter and remove the sweet wort liquid that would eventually be turned into beer. During the mixing process, when so much of the brew’s flavor, color and aroma are determined, the brewers had little insight into process variables like pressure, flow and temperature, and their effect on product quality. Nor could they easily create comparisons between batches.

“Although the brewers were able to identify variances in the product during the quality phase of its operation, they were unable to perform real-time adjustments that would optimize product quality and consistency, which is critical for the brewery to keep up with market demand in their fast-growing business,” says Fred Bossard, president of Aurora.

Aurora designed the control systems for the critical part of the brewing process (raking and filtering) and applied three manufacturing intelligence software products from Rockwell Automation: FactoryTalk Historian to create an accurate online record of production parameters; FactoryTalk View SE for the interface; and FactoryTalk VantagePoint. The latter correlates information captured from the control system, historian, and FDA databases that previously had to be hand combined from separate spreadsheets, enabling the brewer to make decisions that optimize production.
Now the brewer can track how a pressure variance (caused by mash density variances) affects product quality and make inline corrections to improve throughput. But when that same pressure information can be correlated to outlet flow or buffer tank level, he can also better plan and execute machine maintenance activities. “The system [also] enables operators to pass information accurately between shifts with a permanent record that can be relied upon for regulatory compliance purposes,” says Bossard. “The master brewer can watch multiple variables at different phases of the production process all from a single position, rather than patrolling the plant floor.”

As a result of the implementation, Full Sail Brewing was able to increase throughput from four batches on a 12- to14-hour shift, to five batches on a 10-hour shift. “This dramatic increase in output—upwards of 60 percent—has been critical in enabling the brewer to meet nationwide sales targets,” says Bossard. In addition, filtration process improvements have resulted in more efficient water usage at the plant, and reduced the weight of spent grain that needs to be shipped away from the plant, further increasing profits.

“All too often, there is a brick wall between the business and plant floor sides of manufacturing that no one can see through,” says Bossard. Real-time process information is enabling clients like his to break down that wall, he says, helping them to “overcome visibility, migration and integration hurdles so they can improve their operations—and their bottom lines.”

Emerson (www.EmersonProcess.com)

Invensys Operations Management (http://iom.invensys.com)

Rockwell Automation (www.rockwellautomation.com)

Siemens Industry (www.usa.siemens.com)

Yokogawa (www.yokogawa.com)