Now, let's take a look at the energy industry, or the energy and utilities industry. The types of businesses include oil and gas - upstream, midstream and downstream - electric generation and distribution, and water. The types of analytic concerns that these businesses often have include exploration, research, drilling, refining, distribution or pipeline, tankers, storage, water treatment, employee safety, public safety, land rights. They're also concerned with load, electric load, pollution and climate, chemical analysis, smart meters, customer service, and regulatory issues. So, one major oil and gas supplier wanted to improve the complete management of chemical usage and costs, while maintaining asset integrity of its natural gas wells. Their analytic solution collects data from chemical samples, asset information, corrosion data, well production, events and failures from over 48,000 assets across 28 separate locations. This includes over 16 million points of data. As a result, they now have a unified real-time view of well performance and production and are able to manage chemical cost on a global basis and benchmarked across wells, fields, and different vendors. Chevron wanted to improve the performance and availability of its production capabilities. So every two hours, oil and water samples are taken and entered manually through web forms and into Excel, and then import it into an analytic database. Now the system imports process data, flow rates, pump rates, et cetera. They're automatically entered every two minutes. Chemical analysis is done to calculate daily usage and predict yearly production. These prediction capabilities enabled them to optimize downtime and environmental compliance, and reduce refinery downtime and increase availability by two percent, the equivalent of about $35 million per year. National Grid wanted to reduce the operational expense and capital expense through condition monitoring of its substation assets. They created a strategic asset management system to provide a company-wide approach to centralizing data capture analysis and display within a single application. This connects to all of their condition monitoring devices and provides real-time data through a variety of interfaces. This solution saves them an estimated 20 million British pounds annually by reducing transmission failures and extends the life of assets, reduces service cost, and ensures regulatory compliance and reporting. The California Energy Company needed to improve the reliability of the electric grid, and the utilization of energy to meet the state's renewables goals. Their new solution gives them real-time visualization and analysis of 25,000 miles of power lines. They can re-forecast hourly the generation needs based on wind and solar estimates, including real-time alerts for crisis conditions. Since the solution was implemented, there had been no system-wide outages, and it's enabled implementation of 4,000 pricing nodes, up from five, to facilitate cost-effective local market pricing. Also, it's improved the renewables forecast accuracy by 50 percent. Hydro One wanted to determine the most cost-effective replacement strategy for its aging assets. They created a solution to perform geospatial visualization and analysis across 45 million assets comprising $20 billion across 30 different operational technology and information technology data sources. This consistent application of seven key risk factors across all of its assets drives a clear understanding of asset condition and performance. As a result, they're saving a million and a half dollars per year in fewer truck rolls or fewer truck deployments for servicing. They're also able to achieve a more timely replacement of the assets at risk and reduce their failures. Also, they now spend 80 percent of the time analyzing data and 20 percent of the time collecting it, it used to be the reverse. Vestas is a wind turbine company that wanted to improve the precise placement of its wind turbines to affect their performance and useful life along with their energy costs. They started gathering data on 178 different parameters. This incorporates a 10 times improvement in the breadth of data that they used to capture, and now comprises 18 to 24 petabytes of weather data, including temperature, barometric pressure, humidity, precipitation, wind direction, and velocity from the base of the turbine standards all the way up to the tip of the blades. The supercomputing based analytic capabilities they deployed enables them to model 10 square meter grids versus the 27 square kilometer grids that were previously modeled. This has enabled them to reduce wind forecast modeling by 97 percent, from three weeks down to 15 minutes to pinpoint optimal placement of each turbine. The Mexican oil refinery Pemex wanted to reduce the refinery issues and unplanned maintenance. Often, the refineries would go down due to the failure of one or more components. So, they asked their engineers, "How do we know when a piece of equipment is about to fail?" and the engineer said, "Oh, hace ruido" - it makes noise. So, what Pemex did was put sensors all around the refinery to identify vibration, to take baseline readings of vibration, and identify when vibration gets out of spec. This solution enabled them to shift from unplanned to preventative maintenance and has reduced refinery downtime and save 960 hours per year in manual monitoring per refinery. Another utility provider, TransCanada, wanted to improve the efficiency in its operations with its existing gas turbine equipment. They upgraded each of their 1,600 gas turbines with a 100 physical and 300 virtual sensors to capture millions of hours of operating data. By analyzing this data, they were able to improve their output by 5-10 percent, and the efficiency by one to two percent, including lower emissions. Across the 10 thousand households per site, the annual savings was $900,000, and the reduction in unscheduled maintenance resulted in longer continuous operation and lower operating costs. CenterPoint Energy's customers have smart meters. Today they collect data 273 million times per day up from 80,000 times daily. They've added this to a 1,000 other disparate data sources into a single view, which gives them an entire view of their smart grid that can determine the exact location of power issues, precisely where they occurred and respond to them much more quickly. This gives them greater control of energy consumption for their customers. South Australia Water wanted to provide reliable forecasts for demand movement and the supply of clean water. They import weather data and historical customer usage data to analyze the potential demand for all water data collection every 15 minutes. Their analytics solution creates what-if scenarios in 30 minutes, replacing two-to-three weeks of previous manual analysis that used to be done. They're now able to analyze actual versus predictive forecasts of water demand levels and best routes by zone, and they can calculate water demand up to two years into the future.
