24/7 Operations Centers

By Dave Roberts, OSIsoft

The benefits of windpower in reducing dependence on fossil fuels and supplementing the portfolio of energy and utility companies across the globe are no longer speculative. From California to China and the far reaches of the North Sea, wind farms now serve as a significant energy resource and have reached “utility scale” reliability.

Click here to enlarge image

As the scale of windpower expands and reaches upwards of 10 percent to 20 percent of generation capacity in some markets, evidence exists linking advanced plant engineering and operations management as a critical component to the overall reliability and stability of the power generated. As a result, utility-grade 24/7 operations centers are becoming a requirement.

Already, the demands placed on wind plant performance are increasing daily. More emphasis is being placed on modeling the grid interaction and stability as well as turbine performance and asset management. And finally, forecasting capability-in both market and systems operations (control room) planning-is an important lever in accommodating the continued growth of wind generation for power systems.

An example of this 24/7 operations center is the approach from Iberdrola, a world leader in wind energy with operations in more than 20 countries. To manage all of these wind assets around the world, Iberdrola Engineering (Iberinco) has developed a centralized wind operations center, called WindCORE, which enables supervision, 24/7 real-time data acquisition for operations and maintenance and communication with other energy management centers.

With this real-time WindCORE infrastructure, Iberdrola has developed a sophisticated production and maintenance management system. On the asset management side, a real-time historian collects information from more than 100 wind farms around the world in a consistent way and provides insight into changes in asset performance, in real-time. On the maintenance side, the real-time framework becomes a tool in managing both process and costs, ensuring asset health can be assessed without disruption or costly downtime. On the economic side, Iberdrola uses real-time information to quantify “lost” energy associated with maintenance, grid, warranty and weather issues and, ultimately, to prioritize maintenance and deliver continuous improvements in the operating experience.

Today, there are three primary consumers of utility-scale operational data around windpower projects: (1) the project owner/operator, (2) the power purchasing utility and system operator and (3) the wind turbine manufacturer. All three have different but parallel uses of the knowledge and data presented by real-time infrastructure.

The primary user of the performance data of the wind turbines is the wind farm’s owner/operator. Iberdrola is a prime example of a global owner operator (some of their competitors include Florida Power and Light, Endesa, Acciona and-until recent acquisitions-PPMEnergy). These entities typically employ the build/own/operate model, selling power under fixed-price, long-term purchase power agreements (PPAs) to other investor owned utilities, municipalities and other power purchasers. In addition to selling the megawatt hours produced at the wind farm, depending on the market rules all of the environmental attributes are bundled and transacted with the power as renewable energy certificates (RECs). Last but not least, some remuneration schemes are associated with the quality of the energy served to the grid, such as reactive power regulation capabilities of the wind farm.

Power Purchasers

A secondary user of wind farm performance data is the power purchasing entity and the utility/system operator for which the wind farm is grid connected. The nature of the PPAs is typically a “must take” or a “take or pay” type of contract. The PPA will typically specify what kind of information must be provided to the power purchaser and at what frequency. Some examples of major utility wind power purchasers include Southern California Edison, XCEL, HydroQuebec, BPA and TXU. Multiple factors drive the purchase of renewable energy for these utilities, including:

• Renewable Portfolio Standards: These dictate that some percentage of energy or capacity must come from renewable resources by a specified time. Today, there are approximately 13 state RPS as well as a number of federal proposals. A typical RPS requires 10 percent to 20 percent of generation (energy or capacity) to come from renewable resources by 2010 to 2020. Similar schemes exist in the European Union where a target of 20 percent generation from renewable sources has been set for 2020.
• Fuel Cost: In addition to policy mandates such as RPS, many utilities are looking at the fuel cost hedging capacity of windpower. By its very nature, wind farms have a fixed fuel price (zero) for the life of project. This results in highly predictable power costs over the life of the project (power price variability is typically only tied to a consumer price index/inflation adjustment).
• Carbon Risk: Despite the lack of a carbon framework in the U.S. market, many large project/power purchasers are beginning to address the potential for Kyoto-type impacts that may lead to carbon cap and trade or other market-based mechanisms. Acquisition of low-cost windpower today can provide “banked” carbon benefits in anticipation of future carbon and greenhouse gas-related policies and requirements.

System operators and utilities have an obvious interest in gathering real-time data from the wind generation nodes to resolve the PPA clauses that may be associated with a number of issues coming from higher-level grid control. As the share of wind power on the grid increases, system operators become increasingly concerned about stability issues. Gathering production and performance data and transmitting it in real-time to system operators is becoming an important clause of PPAs.

Turbine Manufacturers

Turbine manufacturers are both a large consumer and a source for real-time data. The greatest challenges to the stability and performance of turbines have historically been the tremendous loads put upon the gear boxes. Typical annual failure rates of gearboxes are in the sub-3 percent range. If a complete gearbox repair and retrofit is required, the costs may approach several hundred thousand dollars of equipment cost, parts, labor, lost-energy and downtime. This presents significant warranty exposure on the turbine manufacturer during a wind farm’s first two to five years of operation. Managing the gearbox risk (as well as risk involving other key systems) requires a significant balance sheet and technical capacity.

All of these information consumers have common security concerns. In the North American market, cyber-security guidelines for power plants are governed and managed by the North American Electric Reliability Corporation (NERC). NERC critical infrastructure protection (CIP) standards are beginning to affect system integration among the various classes of data consumers listed above.

There are some common concepts to be considered in building large scale wind systems across each type of consumer. But at the core is the need for an operational center that can monitor output, health and stability in real-time, managing data consumption and presentation across many types of systems. These operational centers not only provide much-needed control, but also become the center for economic analysis and an integration point in an overall generation portfolio.

Information consumers should consider the following to be of equal importance when designing or upgrading their operational centers.

• Integration: The capacity of the infrastructure to interface to a wide variety of supervisory control and data acquisition (SCADA) systems at the wind farms, as well as “off-site” data collection/observations from sources such as weather, power market pricing, forecasting services, aeronautical and so on.
• Analysis: The capacity of the infrastructure to conduct server-level analyses on common performance metrics, as well as deliver event/exception based calculations and knowledge to the appropriate locations/users in a timely manner. Analysis/event information could also be exposed as “services” to ERP type workflows.
• Collaboration: The capacity of the infrastructure to allow all “role-based” users to use a common platform and working environment.

Building an Operational Overview

For owners and operators, the most fundamental tasks for the operational center are to manage assets with greater efficiency and to provide insight into how all of the various assets-including those that are remote-are performing.

Basic functionality and raw data often reside within SCADA systems at the individual wind farms, but that data also has to be available at a high resolution, managing upwards of at least 300 data points per turbine alone, and gathered consistently over time to be useful in a portfolio management schema. This is compounded by the diversity of the various SCADA systems in the asset fleet. It is not uncommon for an owner to have dozens of different SCADA systems within their asset fleet. Combining this data into a common infrastructure and workflow is a critical success factor.

Ideally, the operator can view wind farm information in a rich, high-level overview, with the ability to drill down to information as needed. This may include: geographic location of assets (interactive with map/geographic information system solutions), site and project status (output, wind speed), regional rollups and projected output.

The next level of real-time analysis should be viewed in the context of project-level operational information including turbine map, turbine status and availability, operating efficiency, forecasted output and meteorological information. Substation and meter data including VAR/power factor, phasor/PMU data should also be easily viewed in a real-time infrastructure.

Because the health of the turbines is critical, an individual “zoom screen” view of each turbine should also be available to operators. These drill-downs must contain leading indicators such as all sensor-level data including wind speed, power, pitch and temperature (from SCADA systems), alarms and warnings (from SCADA or condition-based maintenance-CBM-systems), vibration and harmonics data (from CBM) systems, oil analysis from laboratory information management systems (LIMS), as well as turbine availability and efficiency.

Finally, in an effort to manage not only the personnel and subject matter experts available, but also the accumulated knowledge of the facility, an operator may want to view and access data from maintenance logs, work schedules, resource planning personnel rosters and other knowledge management and response resources. This is especially important in managing remote assets, where dispatched technicians may take hours to arrive.

As more emphasis is placed on enterprise-level visibility, many data consumers will also need a way to manage and report on aggregated warranty data. This may include basic production and alarm reports, but should also include a way to audit the data, should it be required in the future. Enterprise visibility may also require integration with a maintenance management system, such as Maximo or SAP’s PM module.

A View of the Future

Iberdrola recently implemented WindCORE (developed by Iberinco, Iberdrola’s Engineering Services firm) to help manage their own advanced windpower operations centers.

WindCORE is a wind operation center designed to meet specific technical and functional requirements for not only Iberdrola, but for Iberdrola Engineering’s (Iberinco) windpower customers around the globe. The WindCORE approach was designed from the onset with advanced development and integration capabilities, which allow easy and secure integration of future applications and functions to comply with evolving, cross-border electrical regulations. As one of the largest application SCADA in the world, the WindCORE system is presently installed and communicating with more than 100 wind farms and supervising more than 1,200,000 points in real-time.

At the core, a SCADA-integrated, real-time calculation database performs periodic average calculations with full control of samples and their quality effects, for “if-then-else” calculations and also algorithms (such as vector treatment for wind speed and direction, calculated values dependent on other calculated values, curve interpolation and so on). Unlike offerings from control center vendors, the WindCORE architecture adapts better to windpower generation-specific characteristics expressed inside the technical requirements:

• Very large, real-time data processing and different information levels supervision
• Multimedia alarm manager for maintenance optimization and, as a result, better wind farm exploitation
• Configuration tools optimization for wind farm repetitive process.

Integral to the WindCORE design approach is a real-time infrastructure. The vast amount of historical information that wind operations centers have to deal with is about one order of magnitude larger than the size of the biggest historical database found at traditional energy control centers. The historical database has to maintain performance with the addition of new tags and also usage is not as clearly defined as in other applications (10-minute averages versus five-minute averages, for example). All these characteristics render historical database selection and structure design critical. Using a time-series, compression-optimized historian capable of handling millions of tags and generating integrations, averages, maximums and minimums on the fly is critical in managing windpower.

The Changing Role of Windpower

Utilities everywhere are tasked with integrating larger percentages of renewable energy into their portfolios. Iberdrola has implemented the WindCORE to help them manage their diverse portfolio of assets. Some of the benefits Iberdrola is already realizing include:

• Greater than 1 percent estimated increase in operation and maintenance cost reduction (and resulting asset availability gains)
• O&M resources optimization and dispatching
• Trading operations optimized through very precise production forecast
• Remote assistance from centralized performance engineering resources
• Fault/lost energy calculations for economic dispatch requirements.

Windpower is among the lowest-cost alternatives to satisfying renewable portfolio standards (RPS) and other policy mechanisms in both U.S. and foreign markets. As the portfolios of these assets get larger, the need to maintain maximum “in market” availability and operating efficiency is critical. Utilities are seeking ways to rapidly and effectively integrate large volumes of intermittent resources, as well as deliver firm green products to the customers they serve.

At the core of an operational strategy to increase windpower within the portfolio is the need for 24/7 operational center driven by high resolution data. A real-time infrastructure delivers a higher level of visibility for operators and where data can be better utilized for analyses, trending, monitoring, and maintenance practices-all of which contribute to full optimization of their assets. This is the critical link to optimizing wind farm assets and realizing incremental gains critical to improving the bottom line. Even a 1 percent increase in availability or output results in significant improvement to an organization’s bottom line.

Author: Dave Roberts joined OSIsoft as Director in May 2005 from SeaWest WindPower. He manages OSIsoft’s vertical industries segment. Mr. Roberts remains active in the wind industry, and works with many of the major wind and power generation companies to optimize the management of their various assets and businesses.

Defining Key Windpower Performance Indicators

• Actual Output: The energy that the turbine has really produced for the actual wind conditions within the measured period.
• Average Wind Speed at Wind Turbine: The average wind speed measured at the turbine over a period of time.
• Budgeted Output: The energy that the turbine or the project should have produced based on the actual wind conditions within the measured period (for example week, month, year). This indicator is calculated from the theoretical turbine or project power curve considering an availability of 100 percent.
• Budgeted Wind Speed: The normal average wind speed over a period of time.
• Capacity Factor: The wind turbine’s actual energy output over the measured period of time divided by the energy output if the machine had operated at its rated power output for this period. No adjustment is made for downtime. The prevailing wind speed and the availability of the turbine affect the actual capacity factor for any given period. Capacity Factor = Actual energy produced / (rating * hours measured).
• Gross Projected Output: The total energy that all or any of the wind turbines in the project should have generated over the measured period of time, based on the average 10-minute wind speeds calculated and recorded with respect to each wind turbine during the respective time period, assuming that each wind turbine performed in accordance with the power curve.

For example, Gross Projected Output can also be expressed as: Gross Projected Output = 100 percent of the energy which should be generated during the respective time period, calculated by summing the energy obtained from the power curve using the 10-minute average wind speeds over the entire respective time period.

Gross Projected Output excludes production losses due to the unavailability of the project during utility or electrical outages or failures not caused by the wind turbines. It also excludes production losses due to maintenance or repair of the electrical infrastructure, grid system, transformers or any portions to which the wind turbines are interconnected.

Gross Projected Output includes production losses due to scheduled and unscheduled service, maintenance and repairs of the wind turbines.

• Operating Efficiency: Actual Output divided by Operating Net Expected Output.
• Operating Net Expected Output: The aggregate energy expected to be produced while the turbines are actually operating, excluding the production losses during turbine downtime.
• Production Variance (Budgeted vs. Actual): The difference between the Budgeted Output and the Actual Output within the measured period.
• Project Availability: The mean availability of all the project turbines within the measured period.
• Turbine Availability: The percentage of time the turbine was available to produce power (for example, when the turbine is not under maintenance or in a faulted condition) within the measured period (excluding the high wind downtime periods). Warranty conditions with the wind turbine’s manufacturer are negotiated using this value; its correct definition and measurement are key business variables.
• Warranty Efficiency: A percentage derived by dividing Actual Output by Gross Projected Output.
• Wind Variance: The wind speed statistical dispersion from the average wind speed over a period of time. -Dave Roberts