Marketing

Challenge:

In the process manufacturing industries, it’s difficult to characterize erosion by just monitoring the univariate process variables around valves. This lack of visibility leads to failure incidents, which are a risk to worker health, plant safety, and the surrounding environment. These incidents can also quickly lead to additional costs due to unplanned maintenance and production downtime. 

Solution:

Using Seeq, manufacturers can implement a condition-based monitoring analysis to monitor valve health across their fleet and accurately estimate time-to-failure. Subject matter experts can utilize historical failure data to create a predictive maintenance forecast to preemptively detect valve performance issues before failures occur. Valve health is determined by monitoring its performance with first principles analytics (valve Cv), which leads to identifying bad actors.

Results:

The likelihood of unexpected valve failure is significantly reduced, saving costs due to lost production and unplanned maintenance. Implementing this analysis will also decrease health, safety, and environmental risks due to loss of containment if any valve fails. 

Data Sources:

• Process Data Historian: OSIsoft PI
• Maintenance Records: SAP

Data Cleansing:

• Use Agile Filter to smooth noisy data
• Remove production downtime periods from analysis data set

Calculations and Conditions:

• Formula to calculate first principles parameters: Operating Cv (actual) versus Theoretical Cv (expected) based on Flow Rate, Valve Position, and Pressure Drop across valve; also calculate Delta Cv
• Boundaries and Value Search to find areas where Operating Cv is > +/- 10% if Theoretical Cv
• Prediction to create a regression model to estimate Delta Cv behavior into future based on past historical failures (training window)

Reporting and Collaboration:

• Prediction monitoring Trend in Organizer Topic 
• Tables showing Estimated Failure Date / Recommended Maintenance Date / Time-to-Failure
   

Challenges

At biopharmaceutical operations, fill weight analytics are essential in ensuring ready-to-ship products. Often while tracking the consistency of fill weights during batch manufacturing, manufacturers only find out if a product is over or under fill weight at the end of a batch, resulting in a product that needs to be scrapped or inadequate manufacturing capacity. Fill weights are critical for product quality assurance, ensuring consistency over time across different products, dosages, batches, and time points. 

Multivariate analysis is important to be able to predict conditions. Analysis on complex and multivariate data is challenging to use for determining and predicting when operating conditions are occurring that could be an indicator of a ‘bad batch’, signaling that proactive intervention can occur.

Monitoring consistency of batches in near real-time, predicting batch quality, and performing root cause analytics to identify and determine the cause of batches outside of optimal operating conditions is difficult and time-consuming in applications such as spreadsheets. They also make it difficult for teams to collaborate and share insights on commonly accessible data and analytics work efforts within work teams, across different manufacturing sites, and across global operational reporting.

On top of everything, it’s additionally challenging to aggregate multiple processes and contextual data sources into a single application. 

Solutions & Benefits

Using Seeq deployed on AWS to notify operators and engineers of conditions empowers teams to predict a future bad batch. Self-service analytics enable subject matter experts (process engineers, managers, and operators) to easily access, cleanse, contextualize, and perform advanced analytics and machine learning on industrial data.
 
Seeq’s analytics tools are broadly applicable, adding value across dozens of use cases, from control system validation to golden profiling to preventative maintenance. The ease of use for process engineers and teams to gain and share insights quickly on time periods of interest within the manufacturing process leads to immediate improvements to operational performance. 

Seeq increases the value of investment in OSIsoft Pi and Pi Asset Framework. Seeq’s extensive support for OSIsoft’s solutions enable teams to develop insights on one asset, batch, or process and quickly scale that analysis to hundreds of assets leveraging their asset framework. Seeq on AWS enables organizations to easily scale up, adding new users to new manufacturing sites and new data sources (including future plans to add Amazon Redshift as well as MES system data) across their global locations. 

Seeq on AWS also ensures application accessibility for remote and distributed teams with access to near real-time data while engineers and teams are working from home or from remote locations. The application provides ease of enterprise software procurement via Seeq on AWS Marketplace.

Data Sources

• OSIsoft PI
• Amazon Redshift
   

Challenge:
Each substation in a transmission and distribution (T&D) system has multiple transformers feeding distribution circuits. Each of these substations has a set capacity, and utilities must report capacity overages to regulators. Generally, the sum of all substation transformer ratings defines a substation's capacity. In this case, a utility could monitor the sum of MVA load for all transformers to total MVA normal and 24-hour emergency ratings.  

In order to avoid capacity overages, T&D operators install batteries. Batteries are expensive, so T&D operators must decide carefully what locations make the most sense. Considerations include:

• Capability of existing transformers to serve present peak loads
• Capability of existing transformers to pick up full load of station when other transformer fails
• Asset health
• Outage impact
• Forecasted demand
• Cost and pricing of battery installation

Solution:
Seeq is used to monitor a signal against thresholds and investigate what would have happened if one transformer had failed during a certain time period. These same analytics are also applied across a field of substations using asset swapping.

Results:
Implementing Seeq advanced analytics reduced Capital expenditures by prioritizing battery installation based on actual equipment condition and risk of overload. It also improved system reliability to avoid $10s of millions in unplanned downtime and capital expenditures for premature replacement.

Seeq’s advanced analytics enable the aggregation of data from multiple data sources, investigation of "what-if" scenarios, and rapid iteration across assets.

Installing batteries at a substation improves grid reliability and the ability to meet demand anytime. Specific benefits include:

• Solar smoothing: smooth short-term changes in voltage due to intermittent generation
• Distribution deferral: non-wires alternatives to defer or eliminate the need for traditional utility upgrade
• Outage management: reduces the cost of deploying mobiles for contingency resources during substation construction
• Microgrids for critical facilities: allows critical facilities to operate independently of the electrical grid during extended grid outages
• Peak reduction: help resolve potential overloads, address power quality issues at host sites, reduce bills for public sector customers
• Energy cost: discharge batteries every day for one to four hours, reduction in the amount of energy to purchase and charge battery at night at a much lower cost

Data Sources:

• Load data is stored in PI
• Capacity ratings are stored in a SQL database

Data Cleansing:

• Use Low Pass Filter to create a smoothed version of the load
• Signals are cleansed to remove outliers, downtime and abnormal operating data before establishing monitoring boundaries and models

Calculations and Conditions:

• Periodic condition
• Formula (splice)
• Deviation search
• Low Pass Filter
• Signal from Condition
• Scorecard
• Asset swap

Reporting and Collaboration:

• Trends, Metric tables, multivariate scatter plots, and Treemap visualizations are combined into an Organizer Topic for quick consumption of the analytics by stakeholders.

Cement producers need to leverage their data to measure manufacturing productivity, ensure product quality, and monitor equipment performance. Advanced analytics can be utilised to gain important insights for maximising production efficiency, which are key for continuous improvement of manufacturing processes.

Data is available at every cement plant, although it is not always easy to access or leverage. Disparate data sets are spread across historians, smaller relational databases, laboratory information systems, asset management systems, computerised maintenance management systems, and other sources. Advanced analytics applications can connect to all of tthese data sources and provide a central location for subject matter experts (SMEs) to access all relevant data.

This article presents three case studies demonstrating how advanced analytics is used to improve process performance, quality, and availability.

Challenge:
Knowing when to replace a substation power transformer is a difficult task. How do you know you have utilized the full life of the transformer? Most transformer manufacturers define the end of life as age 40. However, lightly-loaded transformers may not reach the true age of 40 until their 60th or 70th year in service. Without the correct data analytics tools, the effects of loading beyond a transformer's nameplate rating over time are extremely difficult to calculate, making it almost impossible to know a transformer's true age and when it should be replaced.

Solution:
The IEEE C57.91-2011 standard provides algorithms for calculating the aging acceleration factor for a given load and temperature and the percent loss of life of transformer insulation. Using Seeq, users can combine real-time temperature and loading data from data historians with nameplate/characteristic data stored in SQL databases to calculate true age for a system of transformers.

Results:
The use of Seeq advanced analytics reduces capital expenditures by prioritizing replacement based on actual equipment condition and improves system reliability to avoid $10s of millions in downtime and capital expenditures for premature replacement. The Seeq application also allows for aggregation of data from multiple data sources, extrapolation of data to installation year, and rapid iteration across assets.

Data Sources:

• Load, temperature and oil/gas analysis data is stored in PI
• Equipment heat run properties are stored in a SQL database

Calculations and Conditions:

• Formula
• Value search
• Scatterplot
• Scorecard
• Asset swapping
• Treemap

Reporting and Collaboration:

• Trends, Metric tables, multivariate scatter plots, and Treemap visualizations are combined into an Organizer Topic for quick consumption of the analytics by stakeholders.

Challenge:

In the electric utilities industry, accessing existing data is difficult due to the required data often being distributed across data sources. The data is often incredibly dense, both in frequency and volume, with thousands of assets. As a result, data is often used only for reactive troubleshooting rather than proactive decision making around capital expenditures and the proper operation of assets.​

Solution:​

Seeq allows the end-user or subject matter expert (SME) to unify data access, allowing for a live connection to data across many different databases from a single application interface. The subject matter expert can interact directly with the required data, without having to go through extra steps to ETL and align the data, to perform calculations or condition-based analytics. These analytics are then scalable across assets, leveraging an existing, or built-in-Seeq, Asset Tree and the Asset Swap functionalities and Tree Map view. ​

Results:​

SMEs have a reduced barrier to entry for analytics, empowering them to use existing data to:​

• Reduce OpEx and CapEx expenditures by extending the life of assets and plants through proper, data-driven preventative maintenance ​
• Gain an increased lead time for decisions/maintenance to minimize costs and impact to operations​
• Reduce catastrophic asset failures by increasing visibility to KPIs and early failure detection

Data Sources:​

• EMS/SCADA, DMS, ADMS, AMI, Metering, Relays and Sensors

Data Cleansing:​

• Signals are cleansed to remove outliers, downtime, and abnormal operating data before establishing monitoring boundaries and models.

Calculations and Conditions:​

• Capsules are created for unique modes of operation. Metrics are calculated to analyze the time aged and time elapsed. Calculations are scaled across multiple transformer assets leveraging the Treemap view in Seeq Workbench.

Reporting and Collaboration:​

Trends, Metric tables, multivariate scatter plots, and Treemap visualizations are combined into an Organizer Topic for quick consumption of the analytics by stakeholders. Visualizations are configured with auto-updating and scheduled date ranges so that reports were ready for daily morning meetings.

Challenge:
An in-depth understanding of Overall Equipment Effectiveness (OEE) across a site is critical to identify process bottlenecks and maximize production. OEE analysis can be difficult to standardize and scale across many similar or dissimilar assets that are found at a manufacturing site. It is useful to develop a simple scoring method to categorize various process units on a spectrum of overall effectiveness. 

Solution:
A large-scale manufacturing operation implemented Seeq’s advanced analytics. In Workbench, the teams are now able to utilize the Point and Click tools to identify unique modes of operation and time spent in each mode. Their engineers are empowered to use historical benchmarking to identify appropriate threshold limits to differentiate between ideal and non-ideal equipment operation. The Seeq integration with existing asset hierarchy systems enables analysis to be scaled across all site equipment. 

Results:
Implementing the plant-wide OEE Dashboard, including high-level comparison across process units, unveiled some unexpected bottlenecks. While the site had historically been looking only at uptime as a means of measuring OEE, they were ignoring large periods of time when the unit was running but under some constraint. This analysis across assets enabled them to identify which processing units saw the greatest rate constraints while running, investigate the root cause of those constraints, and invest in those areas by installing capital projects to de-bottleneck the process. 

Data Sources:

• OSIsoft PI + Asset Framework

Data Cleansing:
Seeq capsules were created using Value Search to differentiate between each of the various modes of operation.

Calculations and Conditions:

• Asset Trees
• Asset Swapping
• Treemap
• Histogram
• Value Search
• Signal from Condition
• Scorecard Metric
• Formula

Reporting and Collaboration:
A high-level dashboard was created to showcase the "big picture" at the top of the dashboard. The treemap color-coded by OEE score of the various process units is interactive, and consumers of the report can click into a unit shown in red to gain further insight into what aspects of that unit are driving the low OEE. 
 

Challenge:

Many continuous manufacturing units run into process throughput constraints over the course of a run. These constraints are often reversible but come with the high cost of shutting down to clean or maintain equipment. The manufacturing plant must balance the cost of the shutdown with the regained efficiencies to optimize the overall production rate to meet its targets as soon as possible. Meeting targets sooner translates into more production and increased profits in the long term. Developing solutions to these types of optimization problems can be complex and often requires advanced modeling packages and programming experience.

Solution:

Seeq Formula can be used to calculate the number of shutdowns that minimize the total time required to produce a given order size. Once the number of shutdowns is determined, engineering teams can calculate the length of the run times between shutdowns and create a golden profile of these run cycles. The forecasted profile can then be used to compare against the actual production rate to understand if the operation is on track to meet their best-case order fulfillment date.

Results:

A sold-out production unit has been looking at ways to increase capacity in tiny increments. By implementing this proactive, optimized downtime strategy, they were able to meet supply chain targets an average of 11% sooner over the course of the year. This allowed them to increase production volumes for multiple products, growing sales and market share.

Data Sources:

• Process Data Historian (OSIsoft PI, AspenTech IP21, Honeywell PHD, Wonderware, etc.)

Data Cleansing:

• Invalid data was removed
• Capsules were created for the different production campaigns

Calculations and Conditions:

• Seeq Formula was used to:
  • Identify the current run’s existing data set
  • Create a continuous signal for the time elapsed since the start of the run
  • Create a running production totalizer
  • Calculate the number of downtimes that would be required for a given order quantity
  • Calculate the total cycle time (time between downtimes) for a given order quantity
• Signal from Condition was used to calculate the minimum cycle time for the order size
• A capsule was created from the start of the run to the point of optimum cycle time
• A Reference Profile was built to project optimal cycles for monitoring performance from the best case

Reporting and Collaboration:

The summary of this analysis was added to an Organizer Topic report that is run after the first few days of each product campaign. It’s used to evaluate production rate, degradation rate, and inform operations whether continuous (degrading) operation or periodic shutdown and maintenance will achieve their desired production outcome sooner.

Janice Abel, Principal, ARC Advisory Group, and Morgan Bowling, Sr. Analytics Engineer, Seeq discuss advanced analytics and machine learning (ML) innovations for process manufacturing data.

Michael Risse, VP and CMO for Seeq, and J.P. Bauman, principal at Altira, a venture capital fund that focuses on venture growth equity stage companies serving the energy and broader industrial space, join Keith Larson to share some best practices when buying innovative new technologies.

When we look back at our lives, it becomes clear that all of us have taken a series of risks to get to where we are today. We went out and pursued an education to enable us to succeed in traditionally male dominated fields, often at great cost and with substantial time commitment. We chose jobs that would give us the experience we needed to reach our career goals. Throughout this journey, we have shared our ideas, and there is certainly risk in that, as it can open up one to criticism. Beyond sharing what we think, we also navigate day-to-day risks, such as where we sit in a meeting and when we speak.

Challenge
The plant management team reviews a high level summary report out of plant performance each morning. This report contains data from a variety of different business units and groups (environmental, safety, process engineering, mechanical) across the plant. Each group must manually input their data ahead of the meeting. The management team only sees this report once a day because it is manually compiled by the individual teams. The report is owned by one person, is stored in a place that not everyone can easily access, and is currently generated in Excel with macros that often require manual manipulation. This report must combine data from a variety of different systems and organizations efficiently. 

Solution
Using Seeq’s organizer topic, different areas throughout the plant can collaborate to put together one report. The larger plant operations report is a roll-up of the individual unit reports, maintenance records, and operator shift logs. The report combines the top Key Performance Indicators for each unit with inventory data to get a snapshot of plant operations. Safety, Environmental, and Maintenance data is also included to ensure management can see the full picture of operations. 

Results
A new report was created in Seeq which eliminates the manual data entry and allows members of the team to update the report at any time using the Auto Update function. Everyone can view the summary and drill down to the individual unit reports to view detailed unit operations, loss tracking, and operator shift logs with just one click. The report is stored on the Seeq Server so everyone in the organization can quickly access and update it whenever is convenient for them.

Data Sources

• Process Data—OSISoft Pi or any historian 
• Lab Data – LIMS
• SQL Data – Environmental, safety, and maintenance data 

Calculations and Conditions

• Scorecard Metrics for Key Performance Indicators on each unit – color-coding indicates whether the unit is running to target or not
• Simple Trending of inventories to monitor when tanks are getting low/full for planning groups to monitor
• Treemap to monitor asset wide performance

Reporting and Collaboration
Seeq’s Organizer Topic is used to roll up individual unit reports and auto-update the report on demand.