Predict-It

Highlighted Features

• Advanced Pattern Recognition (APR) Anomaly detection through unique asset models based on prior running history improves availability and reliability.
• Predictive Maintenance Early warning and fault detection moves maintenance practice from reactive to proactive reducing downtime and damage to critical equipment.
• Cost Effective Reduce maintenance and operating expenses. Highest value, lowest cost APR solution available in the industry.
• Highly Intuitive Software models can easily be configured by subject matter experts.
• OSIsoft® PI System® compatibility Leverages commonly used OSIsoft PI tools and software such as ProcessBook®, DataLink®, and Coresight®.
• ECG Support and Services ECG offers monitoring services as an experienced extension of your staff.

Monitor, Detect, and Analyze

Keep an Eye Out
Predict-It tracks trends in process variables on a continuous basis and compares them to the standard for the operation. In this way, the drift of the process variable toward unacceptable areas of operation may be recognized and a potential future outage can be avoided with preventive management steps.

Find the Problem
Powerful algorithms detect subtle changes in equipment behavior days, weeks, and even months before conventional monitoring techniques.

Take Action
The application aids personnel in reviewing potential problem areas and escalating those requiring further action. The robust and accurate models provide reliable information to engineers and maintenance staff for reducing cost, increasing efficiency, and safety.

Predict-It In Action

Proactively Avoid Failures
In a large utility turbine generator, deposits in the stator cooling passages caused gradual local overheating. Post mortem studies found slow drifts in stator cooling water temperatures below alarm setpoint conditions that went unnoticed for over a year. After the initial indications, localized degradation with attendant temperature anomalies continued without alarm until culmination in a catastrophic failure. Trending of these sub-critical conditions against other operational parameters, as done in the Predict-It functionality, would have alarmed a serious condition. This would have prompted an inspection / remedial maintenance prior to what instead became a prolonged, unscheduled forced outage and costly repair of a generator stator bar failure.