In module six of this Industrial Biotechnology MOOC, we're going to discuss Process Economics and Scale-up. And from the perspective of biochemical engineering and chemical engineering that's a course paramount to be able to implement the process. The commercial implementation of industrial bioprocesses is dependent upon achieving an adequate process performance before we start to scale-up. And there are many that have tried to scale-up directly a process, and failed. Because they jumped to the scale-up part without having achieved an adequate performance at the lab scale first of all. Measuring this process performance is of course, important. And that can be simply assessed by using process metrics, which are dependent upon the type of bioprocess. And these process metrics are a quick way to get an assessment of the economics of the process. So we avoid the need to go into detailed economic analysis here by using process metrics instead. The strategy for bioprocess scale-up should be in two parts. First, we need to improve the process intensity by making sure as I show by the green arrow on the plot here, that we achieve sufficient process, metrics at a lab scale, prior to increasing the volume of the process. It's essential, for example, that we get sufficient product concentration before we start to scale up the process, otherwise we will have further difficulties upon scale up. In general, for bioprocesses the scale up itself is not the problematic part. The problematic part is trying to get sufficient process intensity. And that also explains why we put so much emphasis on trying to achieve these metrics at a lab scale. Let's start by discussing a little bit about process competitiveness, and what is required to make any kind of process competitive. It's perhaps obvious to say that the value of a product must of course be higher than the value of the substrate. But it's also very important that the yield of product from the substrate is sufficient, because it's that yield, in combination with the difference in value between the substrate and the product, which is what determines the amount of product we can make in a process. That's a primary requirement for all processes. But, additionally of course, it's also necessary to make sure that we allow for the fact that we need to develop our process very rapidly. We need, also, to account for capital costs, and also for variable production costs. And those variable production costs depend a little on the type of process that we're interested in. There are two fundamentally different types of bioprocess. We can consider first fermentation, which uses growing microbial cells in order to be able to produce a product. In such a process the rate of product formation is dependent upon the rate of cell growth. Because that rate of cell growth may be limited for example upon oxygen transfer, it means of course that we will also limit the rate of product formation. Biocatalysis is where we have the ability now to decouple the growth of the microbial cells from the product formation itself. We can carry that out either with resting cells or with enzymes, but by decoupling the growth of the cells from the reaction itself, we're also able to change the biocatalyst concentration, and that in turn will determine a rate of product formation. So as I indicate here with biocatalysis we have the ability to get far higher values of space time yield. These two types of bioprocess will have different types of metrics which are important. And I'd like to begin by talking about the fermentation metrics. And to do that I've outlined here a very rough fermentation process structure. On the left-hand side of the figure I have indicated a fermenter, which is fed by substrate, but probably other nutrients as well where we grow cells. We follow that with a separation step where we separate out cells from the products that we're interested in. The product is unlikely to be pure, and will come with a number of by-products, and I've called them waste streams here. And this comes from separation number two. This basic fermentation process structure also gives us an indication about what is important from an economic perspective, and therefore what metrics we need to use. For example, the conversion of substrate into product. The yield of product on the substrate, as we normally refer to it as. Like-wise, the concentration going into separation number two will be critical to determine both the size of the separation and how effectively it operates. So let us begin now with the three fermentation metrics which we need in order to be able to assess a new process. Metric #1 is the reaction yield, the fermentation yield, sometimes referred to also as the carbon yield. Essentially how much of the sugar that goes into the fermentation is coming out as a product. That can be expressed as a mole per mole, a gram per gram, perhaps from an engineering perspective kilogram per kilogram. And scientists interested in metabolic engineering after refer to this as a C mol per C mol. It's essential that this carbon yield is high enough relative to the value difference between the product and the main substrate. And the way we determine that is first of all by establishing the theoretical yield that is possible in the fermentation. We need tools to account for some carbon which is required for cell growth, and some carbon will also be required for the maintenance of the cells as well. In other words, for that part, which is independent of cell growth. The carbon which is required for cell growth and maintenance of course, lowers the overall yield of the reaction. Let us take a moment to discuss the influence of yield on the reaction. I've given here four examples where I've listed different prices for glucose as a feedstock and given some different hypothetical yields. If we have yield for example in example one of 0.25, and a glucose cost of $400 per tonne. That gives us a feedstock cost of $1600 a tonne. If the selling price of a product is $1000 per tonne. That means that we lose money in this case. We have a profit margin of -$600 per ton. That's maybe not immediately obvious because if I look at the selling price on the product at $1,000 per ton. And the feedstock of $400 per ton, it looks like we could make money. And this shows the importance of always taking the yield into account. If I increase the yield and also decrease the glucose cost, example two, for the same value of product at $1000 per tonne, we can now make money. Now, $170 per tonne. In example three, I've gone back to the glucose cost but now increased still further the yield. This means, again, that we're able to make money, now $400 per tonne. And in the last example, where the yield is highest, at 0.7. Now again we can make some money, and if I increase the selling price of the product, now to $2,000 per tonne. We can now make $1,430 per tonne as a potential profit. Yield is very important in the process, and always needs to be evaluated for any fermentation. The second metric for fermentation assessment is the productivity of the reaction. Biologists normally refer to this as productivity. Fermentation engineers, chemical engineers refer to this as the space-time yield. But in all cases, it tells us the mass of product per volume per unit time, and has units of grams per liter per hour. For given amount of product, which we need to make in a year. For example, this will determine the size of the fermenters. The total size of fermentation volume that we require. And to put it into perspective, typical values required for a large scale small molecule process are normally greater than two grams per liter per hour. Could also express it as kilograms per cubic meter per hour given the volumes that we're normally dealing with here. Metric number three to assess fermentation processes is the product concentration. Product concentration which leaves the reactor normally in grams per liter. And this determines the operating and the capital cost of the downstream equipment. It will be necessary to concentrate the product and that will take a certain amount of energy, when we leave the fermenter before going into the rest of the downstream process. And that means that a typical value for a large scale process when we leave the fermenter needs to be as high as 50 g/L for a small molecule. That's very far away from what we find in nature, and it requires considerable work in order to get concentrations of this value. Let us turn attention now to biocatalytic processes. I feed my biocatalyst as shown here, in a reactor with a biocatalyst, and also substrate S. I have a separation step in which I separate the bio-catalyst and I recycle it back again to the start of the reactor. This decoupling of growth from the reaction step, enables us to recycle biocatalysts and also to use high biocatalysts concentrations. Later, I need separation number two, where I separate out any unreacted substrate from the product itself. In principle, it should be possible to get very high conversions, up to 100% conversion in these types of reactions, and therefore, it's not always not necessary to have a major separation step of the substrate. Biocatalytic process metrics are based on biocatalyst yield, that's the grams of product to obtain per gram of biocatalyst and as we gain concentration, the grams of product per liter of reactor volume. In this case, the biocatalyst yield becomes important because we need to make sure that all the biocatalyst we put in gives us a return on product concentration. This implies that we are not limited in terms of rate. And it also implies that we have sufficient stability from our biocatalyst, as well. An interesting example, just to go through is the use of an immobilized enzyme, and just look at the costs about the way in which we can consider how much bio-catalyst yield we really require. Listed firstly here, the cost per kilogram of enzyme. And that of course depends on the size of the market first of all, but it also depends on the form of the biocatalyst as well. For a pharmaceutical market, we have an enzyme cost somewhere between $1000 and $10,000 per kilogram. For a bulk chemical market, because it's a much bigger market, we can invest much more in developing the enzyme production process, and therefore the enzyme cost for bulk chemical market could be around $250 per kilogram. That's not the cost of making the enzyme, but the cost which it needs to be sold for. This cost in itself is not what matters. What really matters is how much product we make, and what value product we can get per amount of biocatalyst that we put in. That means that for pharmaceutical processes for example, for an enzyme contribution to be somewhere between $1-$200 per kilogram. The product to enzyme ratio needs to be between 50 and 1,000 grams per gram or kilograms per kilogram as I've shown here. For bulk chemicals we talk about an enzyme contribution as low as five cents per kilogram. And that gives a corresponding product-enzyme ratio of around about 5,000 grams per gram or kilogram per kilogram. Both market size and productivity are of course critical to this calculation. This is for an immobilized enzyme, and I've shown here on this plot, some example of target metrics, again for this immobilized enzyme case. On the left hand axis, I've shown the grams of product we can obtain per gram of immobilized enzyme. This is a log scale and the values vary enormously as we go up the left hand side to values about 10,000 grams of products, the grams of immobile enzyme. On the bottom axis, I indicate the product concentration that we require. In nature, processes operate where I've indicated here at the bottom left-hand corner of this plot, at relatively low masses of product per mass of enzyme, and relatively low product concentrations. In nature, we don't need to have that return, but to make an economic process, we need to achieve minimum threshold values. For a farmer process, we can talk about 100 grams per liter of product concentration. And maybe 100 grams of product per gram of a mobilized enzyme. For fine chemicals, we talk about maybe 200 grams per liter of product. And maybe up with 1,000 grams of product, the ground of immobilized enzyme. And for bulk chemicals, we need to get 300 to 400 grams per liter of product, and 5,000 to 10,000 grams of product per gram of immobilized enzyme. These kind of values seem very far from what we find in nature, but they are necessary to implement a process. And those processes which have been implemented have indeed achieved these kinds of values. A typical kind of methodology which we might consider for thinking about implementing these metrics looks something like this. We begin by defining the added value of the reaction. We look at the value of the substrate, the value of the product, the difference between the two, and we also take into account the yield of the reaction - how many grams of product we get per gram of substrate. We then define some targets for these process metrics for a biocatalyst yield and also for leaving reactor product concentration. These are based upon economic requirements for our process. In the laboratory we can experimentally measure metrics, and we can compare those metrics with the targets, and quantify the gap between the two. Finally, we can define a development strategy based upon the gap in biocatalyst yield and product concentration between what we have in the laboratory, and what we need to achieve to put a commercial process in place. This way of using the metrics to define a development strategy is a very important development. To summarize what we have learned in this module six. Firstly, the product concentration is critically important for these types of processes. Secondly, this can determine the size of the downstream process. And these typical values need to be, for large scale processes, greater than 50 grams per liter for a small molecule. I emphasize here the product concentration because it is one of the most important aspects to emphasize and to get right from the very beginning when we develop a new process.
