First Solar already has highly automated factories and a strong focus on optimization.
The company's vertically integrated continuous manufacturing process and technology are quality controlled under one roof, enabling production of over 50 Gigawatts for 45+ countries. First Solar products are known for having the lowest carbon PV technology and the smallest environmental footprint.
They were looking to launch their flagship product with a new facility, and identify opportunities to ramp up production faster and more efficiently.
We applied the digital twin approach to the construction of the First Solar greenfield factory in Ohio.
The goals were to —
- Understand the feasibility of throughput targets by plant section
- Identify material handling bottlenecks and insights around possible design changes
- Understand the expected utilization of critical tools
- Generate insights on PLC code, MES/PLC integration and AGV system design
The project improved equipment utilization, throughput volume and reduced commissioning time of new assets and control systems by 40%, reducing overall ramp up time and increase production throughput. This digital twin foundation will accelerate future manufacturing startups for First Solar globally.
We began by proving that digital twin for virtual commissioning delivers tangible and continuous benefits with a pilot, setting the stage for a large-scale greenfield project.
01. Simulation of Throughput
We established that the equipment's material handling system, not the equipment itself, was a line constraint that could be addressed with process adjustments.
We looked at two areas of opportunity, equipment utilization of key assets and throughput volume meeting desired daily output. Simulating and visualizing the flow of the system revealed insights about how best to prioritize lines and stage panels. These strategic insights will inform design and drive continuous efficiency for First Solar at this and future facilities.
02. Emulation of New PLC Code
We eliminated 8+ hours of testing & debugging per code update with virtual testing of code.
PLC programs are conventionally tested on the shop floor during physical commissioning, which can cause delays. Commissioning time of new assets and the control system directly impacts ramp up time and production throughput. As code updates are made continuously, the time saved increases with each update, enabling the engineering team to deploy the code and resume production faster. Our work confirmed that virtual testing of code prior to testing on the shop floor minimizes real-world issues and time delays.
The work delivered at the greenfield facility is the largest simulation of its kind. It showcases First Solar's commitment to innovation and pushing the boundaries of technology.
Simulation of the Entire Greenfield Facility
We determined that with simple adjustments the layout design could meet and exceed the desired throughput, taking into account previous equipment performance and production yield.
We looked at two areas of opportunity, Logic to route & distribute panels throughout the entire factory and throughput validation incorporating volume, location & size of buffers, yield and equipment reliability.
Simulation of the full facility delivered exceptional outcomes. The holistic view revealed insights that would be difficult to detect and costly to address in a physical facility, such as the benefits of balancing equipment speed to natural material handling constraints and distributing panels based on unique constraints instead of simple logic. The virtual commissioning approach will inform the design of the greenfield and future facilities, driving continuous improvement of throughput for First Solar.
Digital Twin Adoption*
- Nearly three in four (69%) organizations are already leveraging digital twins
- 71% of those businesses began investing in the technology in just the past year
- Growth continues and mass adoption is expected in the next 2-4 years
A Quick Primer on Digital Twin
Digital twin is a subset of the industrial metaverse defined as a superior representation of a physical entity or system across its lifecycle using comprehensive data, analytics, simulations and emulations.
It enables us to do things that would be difficult or impossible in the physical world:
- Collaborate with teams around the world to make better real-time decisions
- ‘Travel’ to distant or hazardous locations to identify and address issues earlier
- Transport back in time to understand causes to unexplainable events
- Validate new constructions and expansions in a low-risk and low-cost manner
*Gartner Research, Emerging Technologies: Revenue Opportunity Projection of Digital Twins, February 2022
Altair Digital Twin Global Survey Report 2022
First Solar’s digital twin incorporates 4 models to effectively create a virtual replica of their physical process
Virtual models of equipment, robots and conveyers including components such as sensors and motors
Virtual replica of routing logic that incorporates operational disruptions such as downtime, scrap and preventative maintenance
Ability to toggle from simulated logic to real PLC programs that control the virtual model as if it’s the physical process
Accurate virtual representation of equipment and conveyer HMI that allows engineers and operators to interact with the process in a realistic manner
Our work in digital twin created the foundation for future value creation in 3 areas for First Solar
MATERIAL HANDLING & CAPEX VALIDATION
Test material handling systems and routing logic to validate layout design
Test PLC code virtually prior to physical commissioning to identify and address issues
Validate required number and locations of key assets (AGVs, Accumulators)
MES & PLC INTEGRATION TESTING
Test data exchange between MES and PLC(s) prior to physical commissioning
Understand and plan for communication constraints and issues between MES and PLC(s)
Create a virtual testing environment for PLC code changes
Create training scenarios for line operators prior to deployment on the line
Run simulated tests to validate preventative maintenance assumptions and identify opportunities for improvement