Engro Fertilizers
Chemical and Process Engineering Intern · Summer 2016 · Dharki, Sindh, Pakistan
This was my first industrial process work, on an ammonia and urea complex where energy and water are the two largest recurring costs. I modelled the ammonia synthesis loop in Aspen HYSYS and reviewed cooling-tower performance. It is included here because it is where I learned to look for savings in how a plant is operated rather than in what equipment it has, which is the same habit I now bring to water treatment and to program evidence.
The setting
A urea plant is, from a water and energy point of view, a very large heat engine with a chemical purpose. Natural gas becomes hydrogen; hydrogen and nitrogen become ammonia at high pressure over a catalyst; ammonia and carbon dioxide become urea. Every one of those steps rejects heat, and rejecting heat in a hot, arid part of Sindh means evaporating water in cooling towers.
Two numbers dominate the operating cost: the gas burned per tonne of product, and the water drawn to reject the heat. Both are decided far more by daily operating choices than by the equipment on the site.
The ammonia loop and the purge decision
Only a fraction of the hydrogen and nitrogen entering the synthesis converter reacts on each pass, so the unreacted gas is compressed and sent round again. The make-up gas also carries argon and methane, which do not react and do not condense out with the ammonia. Left alone they accumulate in the loop until a large share of what the compressors are moving is inert gas.
The remedy is to continuously throw some of the loop away. That is the purge, and it is a trade with no clean answer. A larger purge keeps the inert level down and the conversion per pass up, but every cubic meter purged carries usable hydrogen with it. A smaller purge saves that hydrogen and spends the saving on compression work.
My task was to build the loop in Aspen HYSYS and find where those two costs balanced under the operating conditions the plant actually ran at, rather than the conditions in the original design case.
How much of the loop to throw awayA textbook representation of the ammonia-loop purge trade-off, not plant data. It shows the shape of the decision, not Engro’s numbers.
5.0%
4.0% of recycle
Argon and methane enter with the make-up gas and do not react. At steady state the amount leaving in the purge equals the amount arriving, so the inert fraction in the loop is inversely proportional to the purge rate. Purge too little and the compressors circulate gas that cannot react; purge too much and usable hydrogen goes out with it. The circle marks the minimum of the summed curve. Constants here are illustrative.
The cooling towers
The second piece of work was a cooling-tower performance review: approach and range against design, fouling on the exchanger side, fan and pump operation, and the cycles of concentration the towers were being run at.
Cycles of concentration is the ratio between the dissolved solids in the circulating water and in the make-up. Running at more cycles means less blowdown to drain and less make-up drawn, but the circulating water gets progressively harder and more corrosive, until scaling or corrosion sets the limit. Where a tower should sit on that scale depends on the make-up chemistry and the treatment program, not on a rule of thumb. The data-center cooling page has an interactive version of the same arithmetic.
The recommendations that came out of the review were operating and maintenance changes, not new equipment: recovering exchanger duty that fouling had taken away, correcting how the towers were being run, and reducing the water drawn for the same heat rejected.
What the work produced
$0.4M+Capital and operating savings attributed to the recommendations
7%Reduction in energy use identified by the cooling-tower performance review
18%Reduction in water use identified by the same review
The employer attributed more than $0.4 million in capital and operating savings, 7% lower energy use, and 18% lower water use to the team’s recommendations. My contribution was the process modeling and cooling-tower analysis described above.
What I carried out of it
The savings are usually in the operation. Nothing on that site needed replacing to use less gas and less water. The gap was between how the plant was designed to run and how it was running that year.operating margin
Most engineering decisions are trades with a minimum, not targets to maximize. The purge rate has no best value in the abstract; it has a best value given gas prices, compressor duty and the inert content of the make-up. I look for that shape in most decisions now, including which evidence is worth buying before funding a program.trade-offs
Water is an energy decision. Almost all the water a plant like this consumes is evaporated to reject heat. Reduce the heat you have to reject, or reject it more efficiently, and the water follows.water and energy
A model is only as good as the operating case you feed it. The design case and the current case differed enough to change the answer. That is the same reason a treatment technology can perform in a laboratory and disappoint in a household.modeling