24 June 2019
During the past few months, our consultants Rik and Haico have worked on a project with their team for a multinational company with manufacturing sites all over the world. At the head office, goals are set to reduce the company’s CO2 footprint by 30% before 2030. The head office tasked all manufacturing locations to come up with plans on how to reduce this CO2 footprint. At the location in France, they wanted some external people to help find solutions to these CO2 reduction goals.
The project team started at the top level; they identified how much energy in total was consumed at the site on a yearly basis. In this case the majority of the energy consumption consisted of electricity and gas.
Since the project was all about CO2, they needed to convert energy into CO2 emissions in order to calculate how many emissions the gas and electricity emitted. In France there is a lot of nuclear electricity which, from a CO2 point of view, is quite a clean energy. It made it clear that the team had to focus on gas. Even though the electricity consumption was significant, they were given the mission to reduce CO2 emissions; not spend less money.
Then, they started to dive deeper into gas consumption; where exactly in the factory do they use gas? It is mainly used to produce steam, but it is used in a lot of different locations in the factory. How much of it is consumed and where?
Like in most factories, there were not enough ways to help them measure this accurately. In an ideal world, they would have a meter for every consumer, but as they are pretty expensive, they are among the things you cut from the budget when you are setting up a plant or process and are running short of budget. But it limits your insights, which made it challenging for the team and they had to find a way to work around this issue. Luckily in terms of data processing, they already had quite a bit of data, and its full potential was not used yet. So, as they wanted to quantify how much steam is consumed per process, they started looking at the process data the Company X had for each process and it turned out they had quite a lot of detailed data (temperatures, pressures, weights, etc.). When extracting that from their system and introducing it into an analyzing tool, Rik, Haico and their team were able to label this data and combine it. Combining such a huge amount of data is something the company actually couldn’t do by itself. Our team looked into the temperature and weight profiles, and by combining the two together we could actually make quite specific calculations about how much steam is processed when in use, and highlight things such as the fact that if the temperature and weights are low, it means the equipment is not in use. But if at the same time there is still steam consumption it means someone did not close a valve – just like if you didn’t close the tap and it is leaking. As these are manual valves, you sometimes have to climb some stairs to find and close them. It is difficult to control all of them.
One of the solutions proposed by Rik, Haico and their team was then to put a signal in order to know when a valve is open and which one it is. Installing this would provide a lot of improvements for them, but would not represent a huge investment.
For the second part of the project, they also looked at the locations that were wasting a lot of heat so that the heat energy used in some places could be reused to benefit some other places in the factory at the same temperature. In the factory it is possible to locate where you have the heat, where there is an abundance of it or where there is a lack of it, and then you check if you can connect them. This had a certain potential for CO2 reduction. At some places water of 50° was flushed down the drain instead of being reused it (for example, another part of the factory was heating water up to 80° while they could have already used the water at 50° from the other part of the factory).
Their boiler is creating steam and if you don’t run it efficiently by keeping the variables stable, it won’t run well. Rik, Haico and their team were not looking in this direction at the beginning because people at the company told them that they were running it fully efficiently and they even showed them some great data. There was however still a gap that they didn’t understand, so they started digging deeper and recalculating everything. In the end, they concluded that the gap had to be in this boiler area. In the end they found out the way the company defined its efficiency was not wrong, but it was a different definition than the one you should use or the one that made the most sense. To them, this was an eye-opener as they thought they were doing it all right, but they had the wrong measurements.
There probably will be a follow-up with them, as the head office is currently deciding which projects from the improvement plans of all plants worldwide to follow up on. The site in France already shared with us that they currently have some great improvements based on the suggestions we gave them.