Why corn syrup is in everything
Corn starch breaks down into glucose, and glucose is not sweet enough to sell. Fructose is. So a corn plant pours glucose through a packed column of tiny beads with an enzyme locked inside them, at about 55 degrees, and the enzyme flips glucose into fructose until the mix sits near 42 percent fructose. That is high-fructose corn syrup: cheaper than sugar, and it pours. US output went from 2.2 million tons a year in 1980 to 9.5 million in 1999. The interesting part is not the sugar. It is that the reactor you size on day one is a different reactor by month two, and most fourth-years size it as if it runs at its ceiling.
Out now
Try Reflux Student free on Aspen Plus
Reflux Student drives your own Aspen Plus V14 from plain English. You type what you want, it opens the case, makes the change in Aspen, runs it, reads the result back and tells you what it verified. Aspen does the math. You keep the judgment. Windows, your own licence, and three free runs to start: it is a trial with a limit on it, not a free product.
Not on Aspen Plus, or on a university licence that will not let you install it? Tell me which simulator you use →
Glucose is not sweet enough
Starch is a chain of glucose. Break the chain and you get glucose syrup, which is cheap, pours well and tastes thin. Measured against table sugar at 1.00, glucose comes in at 0.75. Fructose, the same atoms in a different arrangement, comes in at 1.73. So the whole industry turns on one question: can you rearrange one into the other cheaply enough to be worth it.
One enzyme, locked inside a bead
The answer is glucose isomerase. It is expensive, so nobody dissolves it in the syrup and throws it away: it is immobilised, locked inside porous carrier beads that stay in the column while the sugar flows past. The beads are packed into a steel column, the glucose goes in the top at about 55 degrees, and the enzyme flips glucose into fructose on the way down.
It stops near 42 percent fructose, and that is not a failure of the equipment. The reaction is an equilibrium, and at these temperatures it sits between 42 and 50 percent. A plant that wants more fructose than that does not run the column harder, it separates the product afterwards and blends it back.
The rule most engineers get wrong
An enzyme has a fixed number of working slots. At low sugar most of them sit empty, so adding sugar genuinely adds speed. Past a certain point every slot is busy, the speed flattens, and pouring in more sugar does nothing at all. Double the enzyme and you double the ceiling, because the ceiling is just how many slots there are. Double the sugar and you get almost nothing.
Add sugar, or add enzyme
The plant sits at about 4 times Km. Try to beat the ceiling with sugar.
At four times Km the slots are already 80 percent busy. Doubling the sugar from here buys 11 percent. Doubling the enzyme buys 100 percent.
The bead is never at its ceiling
That is the textbook rule, and a packed bed breaks it immediately. The enzyme is not in the liquid, it is inside a bead, and the sugar has to diffuse in before it finds any. Sugar is consumed on the way, so the concentration falls from the rim toward the centre and the enzyme in the middle of the bead works in a thinner solution than the enzyme at the edge. The whole bead therefore runs below the rate its own kinetics promise, from the first hour of its life.
How far below is set by the bead radius against how fast the enzyme eats compared with how fast sugar can move, which is the Thiele modulus. Small beads are nearly fully used; large beads are mostly dead weight in the middle. Small beads also pack tighter and cost pressure drop, which is the trade.
How much of the bead is actually working
The effectiveness factor for a sphere, with the square root of k over the effective diffusivity taken as 20 per mm.
The sugar never reaches the middle of the bead in any strength, so the enzyme there is paid for and not used.
And the enzyme dies while you watch
Glucose isomerase is run hot because hot is fast, and hot is also what kills it. In one study it lost half its activity in 45 days at 60 degrees. Industrial columns are quoted with longer half-lives, but the shape does not change: it is a first-order decay and it never levels off.
So the operators do the only thing that holds the product spec. If the column has less activity in it, the syrup needs longer in the column to reach 42 percent, which means less flow. They turn the flow down a little, week after week, until the bed is swapped. The composition on the readout never moves. The throughput does, and that is what the decay actually costs.
The turn-down, week by week
Holding the same conversion means holding the same contact time per unit of activity, so the flow you can run is proportional to the activity left.
The spec is held by giving up throughput. Nothing on the product analyser moves, which is exactly why this is easy to miss when you size the thing.
Three layers, and a reactor that changes under you
Kinetics gives you a ceiling. Diffusion inside the bead says how far under that ceiling you actually sit. Decay says how fast that number moves. Size the column on the first one alone and you have sized it for a plant that has never existed, on its best day, which it does not have.
This is the layer I am building: the packed-bed model with the diffusion penalty and the decay clock in it, driving a real simulator rather than a spreadsheet.