How Digital Technology Can Help Lower Carbon Footprint

As owners and operators of fossil fuel plants navigate through the energy transition, digital technology offers the opportunity to perform while transforming, balancing the energy trilemma of providing a secure, sustainable, and equitable energy system each step along the way.

That being said, not just any digital technology applied will result in operations and maintenance (O&M) benefits. You may not achieve your intended goals if the importance of technology adaptability, user experience, and equipment know-how are not addressed from the beginning. Without tailoring the approach to the teams, it may be hard to connect the dots between and among systems. If the solution is too difficult to use, adoption is at risk. And if the expertise of the developers isn’t behind the models, there’s a risk to the accuracy of data.

However, with the right solution, O&M leaders are finding that digital solutions allow them to progress toward a more sustainable future, reducing their plants’ carbon footprints, but not at the expense of cost or reliability. And in many cases, they find that by bringing greater visibility to reliability risks and performance shortfalls, and increasing automation, they can reduce their cost to generate, and improve reliability and availability, while simultaneously reducing greenhouse gas (GHG) emissions.

Operationalizing the Asset with Digital Technology

Starting at the asset level, when industry and equipment knowledge are coupled with AI/ML (artificial intelligence/machine learning) technology, it can be a powerful tool to unlock performance, improve fuel consumption and availability, and reduce heat rate and emissions. Whether applied to a gas or steam plant, such technology delivers real outcomes in O&M.

Gas Turbine Combustion Optimization. Typically gas turbines require seasonal adjustment, tuning, or mapping of flame temperatures and fuel splits to ensure reliable and emissions-compliant operations as weather patterns change by the season. This can be a manual process performed by an expert onsite and more often than not requires an outage, which impacts availability. In addition, manual seasonal tuning is only effective for the precise conditions in which it was completed, and does not enable the gas turbine to efficiently respond to ambient temperature or fuel property changes between tunings.

By utilizing AI/ML technologies to continuously optimize combustion in closed-loop control in place of manual seasonal adjustments, aeroderivative gas turbine operators could achieve the following benefits:

■ 0.5% to 1% reduction in CO₂ emissions/fuel consumption/heat rate.

■ Up to 14% reduction in CO emissions.

■ Up to 12% reduction in NO_x emissions.

■ Improved availability with zero manual tunings or associated downtime.

AI-enabled tuning software that is deployed in a supervisory control system, which is fully bounded by the control system safety-critical programming, can safely use machine learning to find the ideal flame temperatures and fuel splits continually and autonomously for optimal combustion. This is true as critical variables such as ambient conditions and fuel quality change.

Two aeroderivative power plant case studies exemplify the types of benefits such technology can deliver. In the first case study, a peaking combined cycle power plant was subject to fluctuating natural gas composition that caused emissions and operability issues, and required frequent technical intervention, leading to downtime. Remote tuners were often called to make manual adjustments to avoid stage down for high acoustics or blowout, and to address problems with NO_x at baseload operations or CO at low-load operations. With the implementation of AI-enabled combustion optimization, the following results were achieved:

■ CO emissions were reduced by 14% when operating with low specific gravity composition that tends to increase CO emissions.

■ NO_x emissions were reduced by 12% when operating with high specific gravity composition that tends to increase NO_x emissions.

■ Yearly or seasonal tuning events were reduced from four to zero, while avoiding 12 days of downtime.

■ After installation, the site’s high acoustics events were reduced from six to zero during a 12-month period.

The second case study involved a power plant that struggled with emissions to the point that it exhausted NO_x credits one summer, precluding further generation for the remainder of the year. The months of July, August, and September represented one-third of the total typical generation for this site, adding to the criticality of the problem. After a digital solution was deployed to continuously optimize combustion in closed-loop, the following benefits were realized:

■ NO_x emissions were reduced by 10%, precluding the need for a combustion overhaul that would have cost $2 million and resulted in a 12-week outage.

■ Software enabled the site to generate power throughout its high-demand season and beyond without exceeding NO_x credits, generating $300,000 in revenue above the previous season.

■ Yearly or seasonal tuning events were reduced from two to zero, while avoiding six days of downtime.

■ Post installation, the site did not experience high acoustics events.

Steam Plant Boiler Optimization

Traditional, schedule-based control systems for steam plants limit the ability to optimize combustion for heat rate and emissions as the boiler degrades, which can lead to unplanned downtime due to tube ruptures from excessive sootblowing. Several digital solutions have been developed to address these limitations; however, open-looped solutions still leave room for operator inconsistencies, and model-base predictive control has limited effectiveness across the operating range. This can prevent achieving the efficiency desired.

Utilizing a combination of model predictive control and AI/ML technology in a closed-loop system that biases the control setpoints to modify air-fuel mixing and activate blowers when an area of the boiler needs to be cleaned can provide emissions, heat rate, and availability benefits, including:

■ Up to 0.5% reduction in CO₂ emissions/fuel consumption/heat rate.

■ Up to 15% reduction in NO_x emissions.

■ From 10% to 25% improved availability due to reduction in boiler tube leaks caused by sootblowing.

Furthermore, for units with selective catalytic reduction, or SCR, deploying such a solution that optimizes sootblowing and combustion control improves the SCR effectiveness and lowers the cost to operate. It enhances the performance and ammonia efficiency of the SCR by improving the stability and uniformity of the combustion process at baseload or transient operation, while also reducing the amount of primary NO_x to be reduced by ammonia injection. By providing a lower and more-balanced profile, the technology improves SCR removal efficiency, reduces ammonia consumption, and has the potential to reduce ammonia slip.

A closed-loop solution allows for reaching the “best zone” and an increase in adoption, since there is not a high learning curve on how to use the solution. For a three-unit coal-fired steam plant, this technology enabled heat rate improvements of 0.55% to 0.61%, fuel savings of more than $1.7 million per year, and CO₂ reduction of 38,000 tons/year.

Optimizing the Process with Digital Technology

In addition to operationalizing assets with digital technology, operators are finding opportunities in O&M that can impact their carbon footprint and cost to generate by optimizing processes and workflows. Digital solutions are built to provide visibility and drive operational improvements and efficiencies by connecting the data, system, and workflow siloes that hide untapped value. One way to approach this in O&M is by integrating an advanced thermal performance tool within an asset performance management architecture to connect siloed workflows with advanced analytics.

Intelligent Performance. Many fossil fuel plants are operating at levels below their original design, forcing them to cycle more, and to spend more operating hours well below their optimum efficiency levels. This makes it difficult to assess performance shortfalls, risks to reliability and availability, and return on investment from maintenance activities.

1. A digital twin, in this case of a gas turbine, can enable integration of analytic models for components of the power plant that measure asset health, wear and performance with customer defined known performance indicators (KPIs), and business objectives. Courtesy: GE Digital

Heat rate, a measure of plant efficiency, is a moving target that changes with seasonal load profile, operating modes, ambient conditions, and equipment health. Heat rate is not directly measurable in the plant and so without a digital twin (Figure 1) to benchmark performance against, it can be unclear whether a change in heat rate is attributable to a change in operating conditions or a change in equipment health and degradation. Only with clarity of the source and magnitude of the change can the most efficient use of resources be deployed to have the biggest impact on recoverable degradation and ensure optimal performance over time.

As such, traditional thermal performance tools and processes, which have been used for many years, are quickly becoming too cumbersome to meet the speed and flexibility necessary. By not having an advanced performance solution, many plants are operating below their optimum efficiency levels. Furthermore, when asset strategy, performance and reliability monitoring, and maintenance workflows are not fully integrated, manual handoffs prevent system optimization and result in higher costs, more fuel burn, and higher GHG emissions.

If the digital solution is not comprehensive to address these issues, teams are not empowered to make the best decision. It’s not enough to monitor—teams need visibility to the what-if and economic scenarios. Also, the solution’s analytics are only as good as the digital twin—and last, the software only continues to provide value if configured accurately after operational changes.

By combining a flexible, easy-to-use thermal performance application, with the analytic power to deliver a holistic problem-solving solution to maintain fossil fuel plants at optimal efficiency, digital technology can help reduce emissions, heat rate, and O&M costs. Such a solution should include:

■ Advanced thermal modeling and analytics across the entire load range, with actionable recommendations.

■ Analytics for economic predictions, carbon known performance indicators (KPIs), and what-if simulations.

■ Fully integrated data streams into the work management process across the plant or enterprise from asset strategy, monitoring and diagnostics, to maintenance and knowledge management.

2. This illustration details the savings to a power plant operator of early detection of an issue with a condenser. Courtesy: GE Digital

With a comprehensive monitoring and diagnostics application, performance engineers are readily alerted to small but significant changes in operating capabilities early so they can adequately assess and mitigate the issue. In the case of a condenser air leak (Figure 2), early detection from advanced thermal analytics, coupled with the ability to trend performance into the future considering economic impacts, can enable the team to decide whether to pull in an outage to address immediately, or wait, depending on the expected costs and benefits.

Additionally, if it is determined the change is not sufficient to warrant an immediate outage, the application continues to monitor—and if there is an inflection point, and performance degradation accelerates, the team is alerted immediately and can re-assess its decision. When this kind of solution is deployed, detection and resolution of a single condenser issue can result in significant outcomes in O&M, both in terms of cost and carbon footprint.

—Rachel Farr is senior director of Product Management for GE Digital.