Sandia National Laboratories is a multi-program laboratory managed and operated by Sandia Corporation, a wholly owned subsidiary of Lockheed Martin Corporation, for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-AC04-94AL85000.

Data Needs for Benchmarking Your Reliability Performance

Sandia National LaboratoriesContinuous Reliability Enhancement for Wind

(CREW) Project

Valerie Hines, Lead Reliability AnalystAlistair Ogilvie, Project Lead

Cody Bond, Data TeamSAND Report # 2013-6547C

Sandia National LaboratoriesExceptional Service in the National Interest

2

Wind Energy Technologies Department

FOCUS Industry needs Reducing energy cost Promoting large-scale

deployment of clean, affordable energy

GOALS High fidelity modeling Blade design to eliminate

barriers Increased energy capture &

improved efficiency Increased system reliability Testing at reduced cost

CREW: Continuous Reliability Enhancement for Wind

3

Goal: Create a national reliability database of wind plant operatingdata to enable reliability analysis

Protect proprietary information Enable operations and maintenance

cost reduction Increase confidence from financial

sector and policy makers

Benchmark reliability performance Track operating performance at a system-

to-component level Characterize issues and identify technology

improvement opportunities

Sandia partners with Strategic Power Systems (SPS), whose ORAPWind® software collects real-time data from wind plant partners

Key Objectives:

Method:

Performance Dashboard

Cloud based online analysis – 24x7 RAM and Performance data analysis One minute statistical data – everyone else uses 10 minute data ORAP® Transformed data Fault / Event analysis Industry benchmarks IEC / IEEE Availability

reporting NERC GADS reporting Data Completeness

and Quality monitoring metrics

ORAPWind.spsinc.com4

Presentation Content

Preliminary Results for the 2013 CREW Benchmark• Results not yet finalized• October 1: Benchmark published (http://energy.sandia.gov/crewbenchmark)

• Sept. 24-25: Sneak peak at “Optimizing Wind Power O&M”

Illustrations of 3 very different kinds of reliability benchmark metrics and graphs• Spell out the data needs for each

5

General Data Types

Raw SCADA• Key data streams: Operating or fault state, Wind speed, Power• Recorded very rapidly (on the order of seconds)

Summarized SCADA• Raw SCADA that has been statistically summarized

– CREW uses mean, standard deviation, minimum, and maximum; and sometimes mode (most common) for non-numeric values like state

• Recorded at regular intervals (on the order of minutes)– CREW uses industry-standard 10 minutes

(SCADA) Events• Summarized downtime records, with a start and end date• Shows symptom; generally less detail about root cause

Work Orders• Summarized downtime records, with a start and end date• Includes more detail about root cause

Analysis Timeframe (Easy to forget!)• Time period over which data was collected and analyzed

– May be different for different plants or turbines

6

Time AccountingData Needs Raw SCADA

• Select the best ONE state/fault for each moment in time• Map each state/fault to one time category

– IEC’s Availability standard (61400-26-1) provides a great information model• Sum time in each category

Analysis Timeframe• Information Unavailable comes from knowing how much time there was

and subtracting how much time is already accounted

7Event & SCADA Data Source:

Power CurveData Needs Summarized SCADA

• Create discrete “buckets” for wind speed and power, to allow grouping

• Count the number of SCADA time periods in each combination of wind speed bucket and power bucket

• Plot Power bucket vs. Wind Speed bucket, with dot size (or other 3-D option) proportional to the count of SCADA periods

8Event & SCADA Data Source:

Event Frequency vs. DowntimeData Needs SCADA Events

• Map each state/fault to a wind turbine system (or component)• Count total number of events for each system; Sum time for each system

Analysis Timeframe Work Orders

• Combining SCADA with good work orders can reduce “Wind Turbine (Other)”

9Event & SCADA Data Source:

Observations

Even basic benchmarking requires some deepup-front thought • How to organize time (information categories)• How to organize wind turbine (system breakdown)• How to map existing data to desired results

CREW Benchmark results are stabilizing• 2013 Benchmark is looking similar to 2012 Benchmark

Electronics Work Orders are still key • Biggest reliability impact is still from the “Other” system• Automated benchmarking for more detailed root cause

relies heavily on electronic work orders

10

Accessing More Information

The 2012 Benchmark presentation and companion technical report are at http://energy.sandia.gov/crewbenchmark

Sandia keeps an archive of our past wind plant reliability publications athttp://energy.sandia.gov/?page_id=3057#WPR

All wind plant owners, operators and OEM’s are invited to participate.Please contact:

The data in the CREW database is proprietary to our partners. We are not able to disclose non-aggregated data.

• Due to a large volume of requests and limited funding, Sandia is not able to provide customized subsets of aggregated data outside the Department of Energy’s Energy Efficiency and Renewable Energy program.

• Strategic Power Systems, our corporate partner in this effort, may be able to assist with more information about wind plant reliability. For more information, please contact SPS’ Jim Thomas.

11

Jim Thomas, ORAPWind® Project ManagerStrategic Power Systems, Inc.Jim.Thomas@spsinc.com(704) 945-4642

Valerie Hines, Lead Reliability AnalystSandia National Laboratoriesvahines@sandia.gov(925) 294-6490