WITH DIGITIZATION, AUTOMATION, ANALYTICS AND BIG DATA GAINING GROUND, HOW LONG BEFORE SOFTWARE APPLICATIONS ARE CONSIDERED AS IMPORTANT AS THE UNDERLYING TURBOMACHINERY ASSETS?
Two and a half years ago, we devoted a cover story to Big Data (Turbomachinery International Sept/Oct 2013, p.20). At that time, scepticism existed about how much gain could realistically be expected in the turbomachinery field in the face of so much hype about Big Data and the Industrial Internet. Yet far from fading away as a mere marketing ploy, these subjects have grown in vigor and now boast tangible applications.
The recent GE Oil & Gas Annual Meeting (AM) in Florence, Italy, featured several examples of this technology on oil rigs, control rooms, subsea equipment and pipelines. But what makes this topic worthy of another look is the fact that many others in the industry are now developing their own software applications (apps) and applying advanced information technology to the turbomachinery field.
Companies, such as Schneider Electric, SoftInWay, Rockwell Automation, Regal, GE, Siemens, Mitsubishi Hitachi Power Systems (MHPS), and many others are involved in either developing products or implementing projects in the field within this category. Some are taking a proprietary approach to app development while others have opted for an open systems approach where the platform can be shared by everyone. Time will determine which approach will prevail.
In any case, the turbomachinery industry is late to the party. The consumer marketplace is rife with examples of this trend — Uber, Amazon, eBay, to name a few. In the business field, retail, IT and aerospace are in the forefront of this digitization movement.
Jon Harris, Executive Vice President for Technical of BG Group, an integrated natural gas company, heralded the approach of companies, such as Southwest Airlines, which has opted for a simple fleet structure with only one training program. “We need this level of standardation in oil & gas,” he said. “We need to use digital advances in Big Data as a catalyst to building in automation.” (Figure 1).
Figure 1: Companies are developing digital twins for gas turbines that enable more responsive
management and add greater accuracy into maintenance actions
Economic woes
It is not every year that Jeff Immelt, Chairman of GE, visits the company’s annual meeting. But he came this time to speak about digital transformation. Compared to previous briefings on the Industrial Internet (Turbomachinery International Nov/Dec 2014 and Nov/Dec 2015), he played down the hyperbole and the promise, and talked up the fact that digitization is the answer to the industry’s current economic woes.
He began by telling the audience to ignore TV pundits and analysis of the market at times like this. He illustrated the unreliability of most predictions by noting a Goldman Sachs pronouncement from a few years back that claimed oil would reach $150 per barrel and would stay there forever. Now the same analyst firm is saying oil will head down to $20.
Immelt reminded attendees that not so long ago, GE went from selling 200 gas turbines per year to only four due to a market shift. Cycles happen, he said, and the way ahead is to lower costs, become more flexible and improve quality while building your capabilities for the future.
“People are going to be using natural gas for a long, long time,” said Immelt, “and digitization is the way to boost productivity.”
His goals are ambitious: Having 20,000 software developers on the payroll by the end of this year, and about $6 billion in app sales in 2016.
But Immelt added that it is not just about streamlined systems or improved software. The end goal is to provide better pumps, rods and equipment fitted with sensors that facilitate production.
That will take OEMs’ and component manufacturers’ applications that can integrate with the larger system they are a part of. That means every business needs to add the creation of software to their repertoire of essential skillsets.
GE’s mantra is that the physical and digital worlds are converging. With machines and people in better communication, insights can be arrived at in time to make smarter decisions. Immelt advocates standardization as the key to cutting costs. The best way to achieve that is via Big Data, better software, more sensors, increased use of robotics and streamlined production processes.
Accordingly, GE has launched 16 “Brilliant Factory” pilot projects where it showcases virtual manufacturing, 3D printing, sensor-enabled automaton, integration and digitalization of all manufacturing systems.
One example is the company’s Florence turbomachinery factory. The new NovaLT16 gas turbine was developed there in record time using these principles — less than two years to develop as opposed to the norm of five to ten years. Components such as nozzles are being 3D printed in the facility.
“Cost pressure is forcing us to confront the digital side after about a decade of merely talking about it,” said Kishore Sundararajan, Chief Technology Leader at GE Oil & Gas.
His advice to those embarking on this journey is to focus on wherever you feel the most pain. This could be in the areas such as removing downtime, and process or schedule optimization. That provides a specific target for software development that raises the chance of success and a shorter return on investment.
Figure 2: The oil and gas sector is adopting
the latest Big Data applications
GE is attempting to use this philosophy when dealing with customers. Rather than saying, “We have these five products; which one is closest to your needs?” the idea is to find out what the problems are and tailoring a solution to solve them.
On the software side, this means GE has to be an OEM-agnostic and reach beyond high-speed rotating equipment. Whatever the software developed, it has to interact with non-GE turbomachinery as well as static auxiliary equipment around it, as most users source from multiple suppliers.
Real world adoption
Far from being a nice idea, GE has already gotten several projects into operation (Figure 2). The Columbia Pipeline Group, for example, has co-developed a risk management app with GE to assess risk in real time across its 15,000 miles of North American pipelines. The pipeline company was previously dealing in disparate data sets and multiple siloes, which meant jumping from system to system in an attempt to isolate potential threats.
An integrated software algorithm pulls together pig run data, weather, seismic, Geographic Information System (GIS) data, topography (nearby schools, dense population areas and water source data) and so on to more easily calculate risk.
One dashboard lets managers view potential trouble spots and drill down into more detail when they need to. A Red, Yellow and Green system shows at-a-glance status so the company can determine the threat potential of a leak and determine where to deploy their people. Consulting firm Accenture was brought in to assist in system deployment and provide change management with the goal of raising app user rates.
BP, too, is a proponent. Its app monitors a gas compression subsystem that includes different rotating and non-rotating equipment types.
Fawaz Bitar of BP
“BP is collaborating with GE on a pilot called Plant Operations Advisor,” said Fawaz Bitar, Head of Global Operations, BP. “This Big Data project is an example of the kind of innovation that the industry needs, to run assets more safely, reliably and efficiently.”
The tool detects anomalies and brings all relevant data to one screen to help troubleshoot and manage the case to resolution. The basic concept is to treat the system as a whole rather than solely monitoring one component or one compressor.
Ultimately, BP wants to reduce unplanned downtime at it offshore operations by eliminating compressor excursions and leveraging predictive condition monitoring to isolate early warnings of changing facility vulnerability.
In the past, issues were only being detected right before a major failure. Due to lack of time, this required full shutdowns to address issues. By designing the system together with BP, the app will bring the process and control data into one place to provide a more unified view and easier access to data in advance of an event.
“It was found that the control room operators didn’t require another software tool, so we are putting better data from the control side into the hands of process engineers to enable interventions to be planned and executed earlier to minimize process impacts.” said Ashley Haynes-Gaspar, General Manager Software & Services, Oil & Gas Digital Solutions at GE.
This will involve having the control room and process room personnel collaborating closely and using analytics to solve joint issues. BP views the turbine as a component of the larger system and looks at how all related equipment behaves together, including the vapor recovery unit, booster gas compression, export gas compression and gas dehydration system.
These sub-systems are made up of many pieces of machinery, rotating and fixed, both GE and non-GE. The creation of this app will help to close the loop on the entire gas compression subsystem and integrate the efforts of all teams.
“To take the industry to the next level of operational excellence, it is time to take a more proactive approach to collaboration in areas of safety and standardization, while seizing opportunities presented by a new digital age,” said Bitar. “Technology can revolutionize how we drill wells, how we optimize production, and how we detect and react to new information.”
Paradigm is another company working on an app that models and simulates subservice oil reservoirs. “This app required Paradigm’s domain expertise in modelling and GE’s knowledge of heavy equipment to create the desired tool,” said Haynes-Gaspar.
The joint solution is the first to integrate and automate techniques used by production and reservoir engineers so that sub-surface knowledge can be cross-referenced against production data to produce a barrel of oil more cost effectively. By better visualization of the reservoir, it is easier to optimize drilling and extraction efforts.
Beyond the oil & gas sector, GE is touting the benefits of its Digital Power Plant (DPP) program. Based upon its Predix platform, it includes advanced controls, as well as several software apps.
GE provides several apps for the DPP: Asset Performance Management (APM), for example, helps plants reduce downtime and production costs through predictive and physics-based analytics; and Operations Optimization (OO) drives better plant and fleet performance across equipment manufacturers, site configurations and thermal cycles. Since launching the DDP in October 2015, GE has signed 15 customers, including Exelon, RasGas, Bord Gais, Crestwood, PSEG, Sapphire, Engro and Saif.
“Small changes in fuel, maintenance, uptime, productivity, flexibility, safety and labor effectiveness have incredible impacts on overall profitability,” said Ganesh Bell, chief digital officer for GE Power’s Digital Solutions business.
The benefits of the DDP are said to amount to up to $230 million in savings for a new combined-cycle gas power plant. The Digital Wind Farm is also said to deliver up to $100 million incremental value for a new farm (See sidebar).
Figure 3: Siemens is adding greater levels of digitization to its turbomachinery
Digital Siemens
Siemens has its own Digital Enterprise initiative aimed at end-to-end digitalization (Figure 3). The company’s view is that the cornerstones of digital transformation are integrated software tools and systems, industry-capable communication and security solutions as well as data-based services.
An important step in this initiative is its acquisition of CDadapco (Jan/Feb 2016, p.12). This adds industrial simulation tools, such as a software tool for computational fluid dynamics (CFD), which can be added to its Digital Enterprise Software Suite.
Based on the Teamcenter collaboration platform, the plan is to expand this suite to provide seamless integration of PLM (Product Lifecycle Management), MES/MOM (Manufacturing Execution System/Manufacturing Operations Management) and TIA (Totally Integrated Automation).
Siemens is also implementing Big Data analytics and cloud platforms. The Siemens Cloud for Industry, known as MindSphere, provides a way for companies to implement its many software tools without having to erect a computer data center on site. MindSphere is also linked into SAP HANA, one of the top cloud-based analytics platforms.
“Digitization is transforming the entire value chain from product development to operations and service,” said Aymeric Sarrazin, Head of Strategy, Siemens Power Generation Services Division. “Data analytics provides unprecedented performance transparency, creating opportunities to improve asset performance, flexibility and availability thereby leading to improved profitability.”
Specific to the power industry, the company’s digital services are called Siemens Digital Services for Energy, powered by Sinalytics, a secure analytics platform.
It is capable of integrating huge volumes of machine-generated data and combining it with field service data and global fleet performance data, as well as other sources, such as weather data. Over 300,000 devices are connected through the Sinalytics platform, and Sarrazin said over 17,500 software engineers and 160 data scientists are working within this platform.
“For turbomachinery, the vision is to produce digital twins of our assets, which can be scrutinized, analyzed and optimized in real time without the need to move people or assets, thereby saving time and resources,” said Dr. Holger Hackstein, Program Manager, Digitalization, Siemens Power & Gas Division.
Additionally, Siemens is looking to software to help it to implement just-in-time delivery of parts, minimize inventory and ensure that its supply chain and service organizations are seamlessly linked together. The goal is to view statistically as far as three years ahead of time when turbomachinery owners are going to need replacement parts. This is accomplished by developing predictive models of a machine component’s wear over its lifetime.
Another objective is data-driven flexible turbomachinery services. Rather than having a fixed service schedule, contracts offer flexibility in terms of scope of the outage, interval between outages and performance guarantees based on the actual condition of the unit.
When physical service is required, data analysis provides the information to pinpoint the needed repair and ensure the right expertise and materials are available. Performance optimization is part of this, entailing detailed thermodynamic simulation of an entire combined cycle plant.
“This enables us to understand how to run a plant more efficiently by benchmarking its performance against the massive amount of plant and turbine data we possess,” said Sarrazin.
A 3D printing facility in Finspong, Sweden, for example, is using data analysis to repair gas burner tips for Siemens’ SGT- 800 GTs. By using this data-driven printing process, repair times have been shortened from 44 weeks to 4 weeks.
The investment of €21.4 million is part of the company’s plans for the mass manufacture and repair of metal parts with this technology. The project includes Direct Metal Laser Sintering machines from Electro Optical Systems, which cost around $850,000 each.
3D printing allows lattice structures with better heat transfer and fuel mixing, improved coating adhesion and the use of alloys. Siemens has demonstrated lattice structure in blades, fuel strainers and compressor impellers.
“Siemens is looking to develop components faster for the SGT-800, and to shorten repair times from months to weeks,” said Rachel Gordon, Technology Analyst at IDTechEx. “3D printing can be used to shorten design and technology validation, reduce manufacturing and repair time, integrate the design and manufacturing process, reduce cost, regionalize support, and cut spare part inventory.”
Figure 4: Turbomachinery and power plant apps will consolidate data from multiple sources onto one screen viewable on any device
A comprehensive cybersecurity program is also part of Siemens software. This allows confidential data to be transmitted and analyzed securely. Attacks from the Internet can be detected and the organization defended against cyber-attacks on critical infrastructure components.
All products are equipped with a high number of sensors, and give an accurate picture of the current condition of nearly all assets, as well as providing information required for security protection.
“Digitization is the key to merge the available operational data with product lifecycle management, engineering, business data and environmental information,” said Hackstein. “We can now combine domain, context and analytic knowhow as never before.” (Figure 4).
Digital MHPS
Like Siemens, digitization is not new to MHPS. It has been using sensors, digital controls and new diagnostic techniques in existing turbomachinery as well as in its R&D department.
“Production turbomachinery today has more advanced sensors and smart local control capabilities than in years past,” said Mark Bissonnette, Vice President, MHPS Americas.
One example is the auto-tuning system for combustion dynamics and emissions called A-CPFM. This adaptive local digital control uses analytics and sensors along with an expert system to address the changing needs of the power generation industry. Recent enhancements allow real-time adjustments based on gas composition fluctuations and continuous emissions monitoring.
Meanwhile, global monitoring centers focus on machinery reliability through detection of abnormalities with predictive analysis. Centers in Orlando, Florida and Takasago, Japan harness Big Data and analytics to improve the overall competitiveness of the fleet.
Valentine Moroz of SoftInWay
“Increasing digitization of all aspects of the design, equipment manufacture, plant constructability, and operation and maintenance processes are leading to improvements in productivity,” said Bissonnette. “We are installing sensor technology for boiler leak detection, thermal performance deterioration with proposed real-time solutions, as well as performance upgrade verification in consideration of the total plant.”
Beyond the OEMs
A sure sign that a trend is gaining traction is when a larger number of companies get involved. SoftInWay, for example, has been studying how operators and overhaul schedules impact machine longevity.
Data is being gathered from power plant controls systems to build models to help identify the relationship between the original shape of hardware components and subsequent performance in order to improve overall design.
“Various disciplines within the turbomachinery community are merging by leveraging evolving digital technologies,” said Valentine Moroz, Chief Operating Officer, SoftInWay. “Benefits are being felt in terms of machine power, efficiency and economics.”
More and more companies are getting in on the app-collaboration act. Hub City and Marathon Motors (both divisions of Regal) partnered with PARTsolutions to create product configuration software that allows design engineers to create and download perfectly matched gear and motor configurations.
The configurator software is a free download with automatic part number generation, reverse part number lookup, 3D PDF data sheets, 2D drawings, and can be embedded into any existing website or online catalog. Designers can start with the gearbox and find a motor match or vice versa. Configured files are downloadable in 150 CAD formats.
“The configurator creates synergy between these compatible product lines and removes barriers that make ordering difficult,” said Marc Macaluso, Director of Digital Marketing & Services, Regal Power Transmission Solutions.”
In the past, digitization has been limited by the large amount of data, the speed of computer systems and their ability to store data. That has changed “with the advent of cloud computing and the promise of unlimited data at the speed of the Internet, allowing us to conduct analysis with an eye towards looking at assets holistically,” said Tim Pieszchala, Application Consultant, Safety and Turbomachinery, Global Industry Solutions, Schneider Electric. “We can now see the turbine as a piece of equipment in the process as opposed to a stand-alone item.”
By pulling in data from outside the turbine or compressor, people who are developing controllers and software can take into account the influence of valves, heat exchangers, motors and other equipment that are used for the process. In the past, data was available for each individual system but not as a collective whole.
Schneider Electric has developed the Altivar Process, an Ethernet-enabled, variable- speed drive for fluid management with embedded digital services. Pieszchala said that the Altivar Process is the first servicesoriented drive with embedded power measurement and process monitoring & control, asset monitoring and protection, and integration with embedded Ethernet. It can deliver from 0.75 kW to 800 kW. “The turbine or motor-driven package can be integrated with the auxiliaries on the unit such as lube oil motors,” he said.
Rockwell Automation is also embracing digitization as a way to boost productivity. In particular, it is using modelling capabilities for rotating equipment and oil & gas processes. These models run in parallel with actual machines. Model data is compared to machine data to capture any deviation that may result in an unplanned shutdown or drop in productivity.
“Rotating equipment is more susceptible to mechanical failures than non-rotating equipment hence the need for continuous monitoring of this equipment to help ensure productivity and extension of machine life cycle,” said Chidi Egbuna, Global Turbomachinery Controls Lead at Rockwell Automation.
Big Data and analytics allow operators to see things in a different way, such as capturing data patterns to understand what has happened in the past and predict future occurrences. Egbuna thinks that digital advances enable turbomachinery owners to leverage automation to minimize maintenance costs by moving from preventive to predictive maintenance.
Where’s it heading?
In the consumer sector, the popularity of the Apple iPhone has given rise to the App Store with millions of apps offering just about every imaginable function. GE’s vision is that a similar App Store will evolve in the industrial sector (Figure 5).
And the company is putting its money where its mouth is. According to Haynes-Gaspar, GE is investing a billion dollars this year in the Industrial Internet.
“Predix is the operating system (OS) for the Industrial Internet just like iOS is the OS for the iPhone,” she said. “We are going to see the rapid formation of an industrial App Store.”
Apple benefits from millions of developers all over the world. It maintains the platform that all of these programmers use. Similarly, GE intends that others use its Predix platform and that teams of developers will spring up for the industrial sector just like in the consumer sector (Figure 6). To date, however, GE has had to do much of the developing.
“We found that industrial companies prefer that we do the development work,” said Haynes-Gaspar.
Meanwhile, GE Ventures is funding the early stage investment of start-ups that can create apps for the industry. These apps are created using the latest Agile development methodology.
The old way of producing software was to specify the features then hand it over to programmers. About a year later, the application would be implemented. Frequently, it contained bugs and did not represent what the company really needed.
Agile development, on the other hand, works in two or three week sprints after which the progress is shown to the client to ensure it fits their needs, said Haynes- Gaspar. The speed of development is aided by the apps being able to sit on the Predix OS which provides the platform they can plug into and integrate with other apps.
Far from being pie in the digital sky, of the $5 billion GE generated in software revenue last year, Haynes-Gaspar said $420 million of that was in oil & gas. Overall software revenues have doubled since 2011 and are expected to double again by 2020.
Some companies develop proprietary systems while others opt for an open systems approach. SoftInWay AxStream, for example, initially looked at the flowpath of the machine, but developers realized this was not enough.
Figure 6: GE intends Predix to become the main platform for the Industrial Internet, combining IT with operational technology (OT)
So the company considered the whole plant, rotor dynamics, economics and other factors. To do this, it had to develop software that could connect to any other inhouse or commercial system. In addition, the company’s open systems policy means more than 300 customers among OEMs, users, service providers and government agencies have been able to introduce improvements to the software that are available to everyone else using it.
“By marrying software for simulating design, analysis and plant optimization with real data from the field and feedback on economics, it is possible to greatly improve the process and reduce the pain for the owners and operators,” said Moroz. “This aids OEMs with data that could influence future design direction and best practices.”
Rockwell Automation also supports an open approach. “We are focused on the concept of smart assets, which can provide contextualized data to an open architecture,” said Egbuna. “This allows our customers to benefit from apps by Rockwell Automation, our partners, and others within the ecosystem.”
Similarly, MHPS works with other technology companies to implement integrated solutions. One example is the adoption of distributed processing as well as the PI system, which forms the backbone of data collection for many power plants and remote monitoring centers. MHPS control systems securely interface with this system.
“This area of big data and analytics is advancing so rapidly that it is unrealistic to expect that one company can have all the best ideas, so cooperation and open communication among technologists is important,” said Bissonnette.
