Maximizing Efficiency: How to Align Your Metals Production With Energy Consumption Forecast (Part 1)
09 Dec 2022 - Energy, Production
Imagine a metals production plant where you can successfully predict your energy demand and adapt your production plans according to energy availability. To ensure grid stability requires proactive adaptation of the production from planning to execution, following signals from grid operators and from energy providers. This approach will further reduce energy costs and contribute in the transitioning to a pure renewable energy supply.
To predict the energy demand of a plant, it is mandatory to identify the main energy consuming processes. Some processes always consume a constant amount of energy while for others the amount varies. For processes with a constant energy consumption, forecasting is easy. If the process’s execution is planned in a time window, the energy consumption and energy demand is known. If the process can be interrupted (without quality losses), adaptation to energy availability can be easily achieved. Unfortunately, these kind of processes with an energy consumption worth mentioning are rare in steel and metals production.
Typically, the required energy depends heavily on the products and process parameters of the production lines. To model the dependencies, the measured energy consumption of past production by products and production steps have to be analyzed. It requires an Energy Management System (EMS) to record the energy consumption, as well as a Production Execution System (PES) that records the production messages regarding production orders. Assuming that, in a given time, a defined production center needs a predictable energy within small tolerances, combining the data of the two typically not connected systems, allows obtaining the specific energy needed, per production step and line, to produce the product.
The specific energy per product, process, production step and line together with a production planning and scheduling builds the foundation for forecasting energy demands.
Energy Demand Forecasting for Individual Production Lines
The proposed energy-forecasting model operates on time buckets. In the following example, the time bucket has a size of 15 minutes, which is the currently used trading interval size, but can be adapted easily to other time spans. Given that for the production line specific energy consumption for a given product is available and line schedules exist. We can now calculate for each time bucket the energy required if production is executed according to plan.
Figure 1 shows an example forecast for a cold rolling mill “Tandem” in PSImetals virtual factory where the average energy consumption for rolling a specific product is derived by applying a formula based on reduction rate and coil width. Each bar represents a 15-minute interval. The y-axis shows the average power during a time bucket. Colorization is by product family, representing the final sales product, without a relation to the energy demand to produce this product at this rolling step in the production route.
The formula applied here is an oversimplification of the real dependencies. For the demo example, the average power is defined by steel grade (group) applying a factor to the thickness reduction times the coil width, plus a static value that represents the basic power level to operate the tandem mill as seen in figure 2.
A real world model will most likely include additional material and process properties to derive the average power for the rolling process.
Assuming that the energy consumption is purely based on the material transformation, we can express the specific energy as required energy per produced quantity. Here, it is obvious that production speed influences the average power per time bucket. It represents one form of flexibility to adjust average power levels by time interval on the cost of reduced throughput. Figure 3 gives an example of a lookup table used in PSImetals virtual factory to calculate the forecast of a second cold rolling mill.
Energy Demand Forecasting for Production Sites and Whole Plants
The main energy consumers are typically scheduled production lines. However, not all production lines contribute to the same extent in the total energy consumption. Some only reach values that are in the range of the forecast dispersion of the main energy consuming lines. Therefore, it is meaningless to predict the exact expected energy consumption on these lines. These lines can be modeled with a fixed mean energy consumption, during planned production and zero energy consumption during downtimes. In addition, non-scheduled lines and equipment can be modeled as a basic load.
Example of Basic Load
- Cranes
- Compressed air generators
- Climate control of halls
- Office buildings , etc.
These consume energy but only marginally contribute as individuals to the overall energy consumption. This basic load may have seasonal day and time dependencies, which can be expressed in a base load curve.
The total site/plant energy demand forecast is then the sum of all lines plus the base load curve per interval. Figure 4 shows an example of the overall forecast for a PSImetals virtual factory. The colorization is by production line or line group. The height of the color distribution in the bars shows the contribution of the individual production line to the overall energy demand of the site/plant per time bucket.
Flexibility is Essential for a Successful Energy Forecast
Industrial demand-side management in the metals industry depends heavily on identifying flexibility potentials in the production process. Onsite energy production and the implementation of energy storage technologies can contribute to compensate over- and under-consumption from the grid. Flexibility potential in the production process arises from load shifting. To modify the production plan and shift operations on production lines, the whole supply chain as well as the order due-dates have to be taken into account.
The size of the stockyards before production lines, line constraints for scheduling and deviations in energy consumption by product and process determine the extent of flexibility. Competing objectives including reduction of lead times to improve the service level for customers, reduction of stock to reduce capital commitment and maximize production line throughput to reduce production costs have a negative impact this flexibility.
Steadily increasing quality targets and the share of labor costs in the highly competitive steel industry requires minimizing “peak clipping” by slowing down production to stay within given limits of maximum energy constraints.
This forces production planning and control to monitor closely the expected energy consumption and the contractual limits with their suppliers.
Now that you understand the energy consumption process, read our part 2 of this blog post, which details the approaches to adapt production schedules according to energy availability.
Series: Energy Consumption Forecast
- Part 1: Maximizing Efficiency: How to Align Your Metals Production With Energy Consumption Forecast
- Part 2: Maximize Efficiency and Reliably Adapt Production Schedules to Energy Availability
What is your opinion on this topic?
Dr. Stefan Albers
Business Consultant, PSI Metals
salbers@psi.de
After 4 years as Head of Production Planning and Scheduling in a cold rolling mill, Stefan joined PSI Metals in 2007 as a Software Engineer and Project Manager. Due to his enormous experience and expertise, he became a Consultant and Solution Architect. Stefan has exceled in this role, becoming a global and prominent voice in planning and scheduling as well as overall customer business process where he has contributed to various projects worldwide.
Follow Us on
