Customized Startup Procedures & Commissioning/Completions Software
For example, a startup team might check digital valves to make sure that when a signal is transmitted to them to open, they, in fact, open. The team may also check to make sure all bolts are torqued to the proper torque, conduct motor meggers on motors, make sure transformers are operating correctly, and make sure that motor control centers are sending out proper signals to the connected equipment. The startup teams check to make sure everything is working properly before someone signs off on the construction and flips the switch to turn it on.
- A more than 40,000-mile electricity transmission network ? the largest in the nation
- More 765-kilovolt extra-high voltage transmission lines than all other U.S. transmission systems combined
- Approximately 31,000 megawatts of generating capacity
AEP?s Original System: Excel + Email
Prior to the Project Startup Metrics database system built by Terraine, AEP captured completion metrics using Excel, and disseminated these details to stakeholders via email. Device lists (cables, motors, valves, instruments, and IO loops) were saved to individual worksheets within a spreadsheet. Systems information and scheduled completion dates were also saved to some worksheets. As devices were completed, the ?completed? column would be marked for the completed devices, and this information would then roll up into calculated fields which would calculate the percent complete for each system. Charts and graphs would then also have to be updated to display the actual completions as compared to the scheduled completions.
Understanding these elaborate spreadsheets?let alone maintaining them up to date?was a somewhat tedious task assigned to a dedicated professional within AEP. The process was complex enough such that only one individual at AEP was allowed to modify the spreadsheets, for fear of editing something incorrectly and thus messing up the entire system.
Developing a Better Commissioning System for AEP
At the request of AEP, Terraine analyzed their commissioning procedures workflow in an effort to improve their system through use of database software technology. After a thorough understanding of their system was made, visual mockups were created and presented to the management team at AEP. After some tweaks and several iterations, work began on the backend database schema and front-end UI. Four months later, a production-level database system was deployed for their use.
The Completed System
The system designed, built, deployed, and hosted by Terraine follows a very simple logic flow:
- Project, system, and device data is imported into the system via the admin panel.
- As devices are completed, these are marked as completed from the data screen. If a device needs to be retested, the ?retest required? checkbox is checked.
- Punchlists and variances are added when an issue arises, and cleared when the issue is resolved.
- Reporting is as simple as clicking on the reports button and viewing charts, graphs, and lists of tables.
When the logic flow is simple, it is understandable by all stakeholders, including the field techs responsible for collecting the data. It is this simplicity that leads to a successfully adopted system. We are proud of our effort in simplifying a previously labor-intensive and tedious commissioning procedures system for one of the largest electric utilities in the United States.
In 2008, Duke broke ground on a $3.5 billion Integrated Gasification Combined Cycle (IGCC) power plant in Edwardsport, Indiana. At that time, Duke?s commissioning procedures system was a paper-based system, comprised of multiple paper forms completed by hand.
Duke?s Original Commissioning Workflow
Prior to the system we built for Duke, when a Duke power plant was built, all of the commissioning information was collected on paper forms by the field techs commissioning the devices at the plant. Then, punch lists of issues were generated for the construction crews, repairs were made, and several layers of signatures were completed before a device was considered ?complete?. Whenever percent complete metrics were needed, the Duke Startup Manager would estimate percent complete at the system unit or engineering area level based on the paper forms that were submitted and from conversations with his crew. The information was considered qualitative as opposed to quantitative. In other words, these were best guesses.
Developing a Better Commissioning Workflow for Duke
The startup manager for the Edwardsport startup and commissioning project spearheaded an initiative at Duke to improve and essentially digitize the commissioning procedures process at the plant. Terraine was chosen to build Duke?s commissioning procedures software application. Terraine successfully built and implemented this application at the Edwardsport location and was subsequently chosen to develop a similar application for a major plant upgrade project at Cayuga Station, Indiana.
Terraine began by analyzing all the paper forms related to Duke?s commissioning procedures process. After achieving a thorough understanding of Duke?s commissioning procedures, Terraine staff initiated initial backend database work. Terraine worked closely with Duke staff to develop a database system that perfectly matched the needs of their organization. Once a working prototype was completed and populated with equipment lists, users list, startup codes, and other static lists, Terraine deployed the system on a few ruggedized devices for testing on site in a production environment. This led to additional changes, and eventually, the system was deployed onto 60 ruggedized tablet devices, in order to commission all of the 60,000 devices at the $3.5 billion construction project, which at its peak employed over 2,500 construction workers.
Fully Digital Commissioning Checklist Forms
The system designed, built, deployed, and hosted by Terraine handled every single aspect of commissioning at the plant. This consisted of complete digitization of every single type of device used at the plant. Whenever a particular device was needed by a startup technician, the proper form was automatically displayed for the device for data entry. Each device type contained percent complete calculations that were based on the insertion of values into certain database fields, meaning that each time something was filled out on a form, backend calculations were completed to determine the percent complete for that device, for the engineering area, for the startup code, for the system type (electrical or mechanical) and for the entire system. This information was pushed to a password-protected dynamic website, so that stakeholders could, with the click of a mouse, determine percent complete at the plant level, the startup code / engineering area level, and at the device level. This information, along with discrepancy (punchlist) reports, was also disseminated automatically each morning via email delivery to respective stakeholders. In addition, a graph within the web application was used for plotting projected completions against actual completions, giving managers an easy way to interpret how behind or on schedule their team was in nearly real-time.
Automated PDF Generation and Quick Turnover Package Handoff
Upon 100% completion of each startup code, handoff was performed automatically by the system at the click of a few buttons from the document control area of the system. The completed forms were produced by the system in PDF format, which even included signatures captured on the tablets, and the ability for document control personnel was made available for downloading complete PDF turnover packages for further integration into other Duke IT systems.
Award Partial % Complete Credit Automatically
At the direction of Duke, an algorithm was created that would assign a certain percent complete for specific milestones completed for each device. For example, motor meggering checks typically take longer than other simple checks. So for devices containing motors, once the motor megger tests were completed, a value of 40% complete was assigned for that device. Similarly, once a field technician signature was added to the digital record, the device would be assigned a 90% complete value. And once the supervisor signature was added, the device would be assigned a 100% complete value. Validation checks were inserted in between each milestone such that certain milestones and required data could not be bypassed or skipped.
All of these calculations were rolled up from an individual device, up to a startup code/engineering area, and up to the project level. By assigning partial % complete values to devices (and not just either 0% of 100%), a more accurate representation of % complete was achieved for the Edwardsport IN IGCC plant commissioning project.
I was able to determine % complete at various levels at our Edwardsport IGCC plant, and we saved over 500,000 sheets of paper while never losing a single record.
– Chris Green, Former Startup Manager, Duke Energy
A stack of commissioning forms taller than a skyscraper?
The application we built for Duke was responsible for saving 500,000 individual pieces of paper. Stacked together, this amount of paper would be higher than a 16 story building.
Not only is a paper based system costly and harmful on the environment, it is also inefficient, prone to an abundance of transcription errors, and usually contains incomplete data.
Don?t guess percent complete.
As a project or plant manager your most important metric will probably be percent complete. Traditional commissioning systems do not provide managers with the accurate information necessary to calculate this metric. Paper based systems take days if not weeks for information to be transferred from the physical site to upper management. The system that Terraine built for Duke automatically calculated percent complete every four hours, with up-to-the-minute real time data.
Get the whole picture.
The system we built for Duke contained an easy to use interface to pull detailed analytics on every part of the commissioning procedures process. The Terraine platform automatically calculated the percent complete for each department, site, and project, which allowed the project manager to identify areas of inefficiency and act accordingly.
Don?t make important decisions based on inaccurate data.
Traditional commissioning systems almost always contain incomplete and inaccurate data. During the process of transferring 500,000 sheets of paper to excel there will be an abundance of transcription errors. leaving data that is both inaccurate and incomplete. The system we built for Duke eliminated the need for transcribing documents and information. In addition, our system contained built in features that required employees to fill in every part of a form and automatically double checked for outliers in the data. According to Duke?s startup manager at that plant during this project, not a single record was lost by our system.
XForms was designed from the ground up with Cx use cases in mind
Automatically calculates partial % complete
- Assigns 0% to forms not started, 50% to forms partially completed and saved as draft, and 100% to submitted forms
- Automatically calculates % complete at the device level regardless of # of forms assigned to a device
- Automatically rolls up % complete to various groupings and hierarchies
Mobile app features designed to speed up field data entry
- Collapsible headers to mark entire sections of a form as “N/A” (bypass validation controls in a collapsed section)
- Excel-like table grids that work even on tiny smartphones
- Linked fields: mark one field with a value, and all linked fields take that same value initially
- Draw on top of pictures
Works natively on all modern devices
- Android tablet
- Android phone
- Modern web browsers (Chrome, Firefox, Safari, and MS Edge)