For a business to be viable, it must make money. Basic math shows us that the margin to be made on a final product equates to the difference between manufacturing costs and selling price. If costs go up, and nothing else changes, margins (and profits) go down. Industries such as food, beverage and dairy often feel this pressure as markets become increasingly competitive.

For example, the average farmgate price per litre of milk in the UK is a little lower than 30p. A supermarket will purchase the milk from the dairy for around 65p per litre. That leaves 35p per litre to cover manufacturing costs and product margin for the dairy.  So, if the cost of raw materials is fixed or rising and competitive markets are keeping final sale prices low, how can businesses compete? The answer often lies in utilities. The costs of energy and water are increasing and represent a sizable percentage of overall manufacturing costs. If their usage can be optimised, valuable margin percentage points can be gained.

This is not based on major investments and changes, nor is this about
Overall Equipment Effectiveness (OEE)—although this is an important part of plant design. This is simply looking at saving energy and utilities through better understanding of overall plant operations. With the right knowhow, its not as hard as you may think. Many manufacturers are embracing technology that provides better tracking and reporting of energy and utility consumption to improve existing operations and reduce costs—but what is the starting point in finding how to save money on utilities?

STARTING THE PROCESS OF OPTIMISING ENERGY AND UTILITY CONSUMPTION

Ultimately, the aim is to know how much energy and other utilities are used to make each unit of product. By understanding this, trends or peculiarities that may point to areas of improvement can be seen. But how is this information obtained? The data that relates not just to a machine but to specific product types or recipes is required. In dairy, food and beverage industries, a single processing line is often designed to produce differing products, through a change in recipe or process parameters. For a milk producer, for example, one batch may comprise organic milk, another ordinary drinking milk, and a third UHT-processed milk. Each batch type may use different amounts of energy or water, and we need to understand this complete picture.

It soon becomes clear that a critical starting point for energy and utility optimisation is a review of process and instrumentation drawings (P&ID) to understand the overall production process flow. This provides the basis to stipulate what reports are required and what data can be compared. From here, the addition of flow, energy and unit measurement can be integrated into the plant. Of course, there is no need for this process to be applied to the whole plant. A project could start with a single processing line with a plan to phase other lines in later. It may also be useful to assess uncollectable data. For example, seal water on pumps is often lost to the factory floor and cannot be easily measured. However, as it is usually a constant flow rate, a value can be assigned to it to give a more accurate picture of genuine production water consumption volumes against water written off to processes.

Advances in measurement instrumentation have led to a significant reduction in size and cost of these units. They are now small devices which can be easily fitted into existing control panels and readily integrated into either an existing control systems or an I/O fieldbus data collector. Reporting tools also need not be an expensive addition. Existing site reporting software, a Supervisory Control and Data Acquisition (SCADA) capable of trending or even a simple spreadsheet tool can be used to collect and display information about the system in a way that delivers valuable insight to operating efficiencies.

Process specialists that have experience in the types of systems used within a site often know which data is most valuable and important to an optimisation project. Their understanding of how a machine operates, the applications for which it is being used and the changes that are possible to improve efficiency, can help realise significant savings in shorter time periods. They also offer experience in instrument selecion— an area which has many potential pitfalls. Every instrument vendor is likely to position themselves as a one stop shop with products that will meet plant requirements. However, these claims may not inform customers that they do not need the level of accuracy or exotic fieldbus included within an instrument and a much more cost-effective option will be equally effectual. Likewise, data collection using a Programmable Logic Controller (PLC), a machine with advanced capabilities that is designed for full-scale control of operations, may be equally well serviced by a modern unit designed purely for data collection, of which there are many readily available in the marketplace.

once data collection points have been established, instruments selected, and reporting tools identified, the next—and perhaps the hardest—part of the project is how to review, interpret, understand and take action with the data collected. It is not uncommon for users to feel a little overwhelmed by the information the new data collection points deliver; they may not fully understand what it means or see unexpected or unexplained points within it. Indeed, although data science is not a new concept, to the uninitiated, it can just become a cacophony of numbers. The data from the factory floor is so specific to a particular machine, interpretation of the numbers can be difficult without the expertise and experience of what they mean. Reviewing data requires experience but it also requires logical data sets that can be properly compared and understood.

WHY YOU NEED GRANULAR DATA

Lets look at an example of a customer in the food industry that was experiencing an increase in energy consumption for the same month year on year. The logic was good, data verified and production volumes very similar, and yet there was significant increase in energy usage. Of course, seasonal ambient temperature variations can mean products require more heating or cooling, but the example was comparing the same months of the year and this variation should have been negligible. So wher was the energy going? What had changed to create such a notable increase in energy costs?

By reviewing production data and comparing it to the energy data, it quickly became apparent which specific parts of the plant were using more energy. In this example, the culprit was the Clean in Place (CIP) system. There had been more CIP cycles for the same volume of product over the month time span. Why? The answer lay in the order of some product runs. For example, organic products cannot follow ordinary products without a CIP run in between. However, the reverse is not true as no cleaning cycle is required following organic production to ordinary production—after all, there is no such thing as non-organic foods. A modification to the production schedule and re-ordering of product runs saw immediate energy savings and reduction in production times with fewer CIP runs.

This example shows the need for granularity of data and the need to match it with operations. By installing a system with the necessary data collection points and displaying it in the correct way, it became clear wher the issue lay. The first point in this analysis, however, was having the data to eliminate what was not causing the increased energy consumption: changes in production volumes, product types or ambient temperatures. The second was identifying exactly which part of the process was consuming the extra energy. From here, it was easy to see why the extra CIP cycles were occurring.

SUMMARY

Process and system experts can partner with food and beverage producers to help them capture and use information from around the plant. Their knowledge and experience can save money in the cost of achieving a useful utility monitoring system and help identify practical actions to optimise consumption and reduce costs.

A single measurement point is not often useful in isolation. Knowing that the energy bill has gone up is not much use unless we can see trends and compare data year on year, batch by batch, line by line or machine by machine. This gives a holistic view of what is really happening and enables data to be turned into information and actions that can realise significant savings to production costs, and so improvement in the lifeblood of typical manufacturing businesses: profit.