Q1: Hello Mr. Lipták. I read your Jan. 23, 2019, online article, "Controlling the Panama Canal." It was excellently presented. I'm hoping we might enter into a dialogue.
I represent a Canadian engineering group looking at helping the Panama Canal Authority to better control their water management issues.
Do you find the canal presently has good instrumentation, or should the system be replaced?
Sean Murdock
[email protected]
A1: I understand that the locks are planned to be expanded again by 2030, which would give you plenty of time to study the operation and design a safer (cyber-secure) and more optimized control system. The old control room was rather disorganized (it appeared more like a stock market status display room), serving a semi-manual, visual-based operation. When the canal was expanded in 2016, the design of its new control room started to look more like a 20th century design. I haven't seen any sensors, but I assume they're good quality, Chinese instruments—but usually without sufficient backup, redundancy or hacking protection.
The vessels are raised or lowered some 85 feet through the three chambers, each on the Atlantic and Pacific sides of the Gatun locks. When the canal was built, the Gatun dam was the world's largest earth dam, and Gatun Lake was the world's largest artificial lake. As this lake supplies the water to transport some 14,000 ships yearly (plus the water for 2 million people), and climate change has reduced the annual availability of water from 5.2 billion cubic meters to about 3 billion, the key goal of your proposal should be water conservation.
One conservation tool is using reservoirs, which are filled when a ship is lowered, instead of draining directly into the ocean. Several reservoirs are located at different elevations, and their water is used to lift ships going the other way. This water conservation method has presently cut water use in half, and if optimized (if the displacements of the two ships moving in opposite directions are close), then conservation could rise to 75%.
The other method of reducing water consumption is to make sure each lock is completely filled before lifting starts. I've seen water waste when half-empty locks were lifting small ships. Another goal of optimization should be to maximize the ratio of the displacement volume of the lifted ships versus the amount of water in the lock. Your proposal should therefore include both algorithms and alarms to maximize conservation.
Still another key process control task is to make the controls totally proof against hacks and cyber-terrorism. This is serious stuff. Hacking and cyber-terrorism are on the rise, and with over 5% of world trade moving through this canal, it must be protected. Just imagine what would happen if one of these giant ships failed to stop when entering a lock and broke through the gates.
In your case, I'd visit the canal and talk to the process control engineers before preparing your proposal because there must be other possible means of improvement besides the ones I've mentioned. A last resort would be to supplement the water supply with desalinated sea water.
Béla Lipták
[email protected]
Q1: Is it possible to use a stainless steel (SS) 321 orifice plate in medium or high-pressure oxygen service (more than 11 to 42 barg) with a pipe also made of SS 321?
I ask because, according to CGA G.4-4, we can't use a SS 316 or SS 321 orifice plate for over 25 barg oxygen service. Could you clear up this subject?
Mehdi Manouchehri
[email protected]
A1: There's nothing in the flow handbooks or the standards to limit the use of materials for orifice plates. Some calculation programs or procedures have options for entering alloy data to compensate for thermal expansion. Strength is rarely an issue except for flow-limiting orifice plates with very high differential pressure.
Cullen Langford
[email protected]
A2: I found a scientific paper indicating why you should not use SS 316 or 321 in oxygen service at high pressures. It indicates that the SS can combust at high pressures. The authors recommend higher-nickel-content steel for use with oxygen. Many variables will need to be analyzed to select the correct materials.
After consulting the specifications of 316 and 321 stainless steel at the website MakeItFrom.com, I found that using SS 321 would have the following advantages: better elongation factors, slightly higher Brinell hardness and better thermal properties. The only issue here is that, since it has a hardness level higher than SS 316, you need to make sure the flanges and orifice plate aren't subjected to undue pressures or vibration. Also, since the whole system would be SS 321, no galvanic protection would be needed. Finally, the equipment using SS 321 may have a potentially higher price and time delay on deliveries.
Alejandro (Alex) Varga
[email protected]
Q3: Given the following food court consumption per week, how can one calculate the required minimum and maximum stock levels?
Buns:
- Minimum: 400 buns
- Maximum: 1,300 buns
- Normal: 700 buns
Lead times:
- Normal: four weeks
- Minimum: three weeks
- Maximum: six weeks
- Re-order size: 4,000 buns
Monicah Nyambura
[email protected]
A1: This looks like an enjoyable stochastic optimization exercise: determine the trigger value (of bun inventory, or perhaps some other marker) to reorder buns. Optimization seeks the best marker for the decision. It's a stochastic optimization because two key influences (consumption and delivery time) have uncontrolled variation.
I recently published a similar study for controller recalibration, (https://doi.org/10.1016/j.compchemeng.2020.106901) and my optimization book with companion website would also provide relevant reading.
However, I think there is too much missing information to start such an effort. Here are some sample questions that need to be considered for a complete solution:
- Do the buns go bad in inventory (dry out, mold, get squished, etc)?
- What's the distribution of consumption, and what are the delivery times associated with the average and min/max values?
- What's the penalty for not having enough inventory, and is there a penalty for too much inventory? (Both of these questions are related to what the enterprise is all about. If military, for instance, waste may be inconsequential, and all that matters is having adequate food inventory. If this is a monopoly business, then running out of food is just a loss of those sales, but not a transfer of business to another vendor—future sales are guaranteed. If a competition business, then loss of inventory means customers switch to another vendor—both present and future sales are lost.)
- Does the distribution of consumption and delivery time vary with season? (If so, then perhaps the trigger to reorder should also vary with season.)
Russell Rhinehart
[email protected]
