Sunday, 26 April 2015

Totally Total Process Plant Design

I was reading a PhD thesis by one of Mecklenburgh's students (on a computer program to lay out plant) as part of my research for updating Mecklenburgh's Process Plant Layout earlier in the week.

The student starts his explanation by splitting process plant design into process design and plant design, and I was reminded again of Pugh's Total Design.

It is hard to use the word "holistic" without feeling a little bit like an alternative medicine salesman, who also tend to describe science as "reductionist". Unfortunately, the primary criticisms I would make of how process plant design is taught require the use of these concepts, but we should bear in mind that there is no alternative engineering.

So we might split process plant design into process design and plant design. We might split these further and further until we have produced versions of process and plant design which are mathematical problems. Even when we do this, the final problems are very complex indeed, even as straight maths problems.

To take the example of getting computers to lay out process equipment in space, these has essentially been no progress since the 1980s. That PhD thesis claimed to describe a fully functional computer programme for plant layout in the 1990s, but the most recent reviews of the literature make clear that no-one has come up with an algorithm as good as a professional engineer even to the simplified problem they are trying to solve.

This intractable problem is not how to best lay out plant at all. It is how to allocate the arrangement in planar space of objects with simplified characteristics in order to minimize the cost of materials transport between them.

Safety, operability, process robustness, and most cost considerations are removed in order to simplify to the point where engineering has become maths. The space in which this exercise happens is perfectly flat, and adding a second floor seems to make the intractable impossible.

Whilst this has kept academics who think they are working on layout issues busy for thirty years or so, it has not been of the slightest use to professional engineers as far as I know. There were few takers for that PhD students program, despite his claims of utility.

Alternatively we might consider what has happened to process design over the same period. Academics now think that process design is done in simulation programs and optimised with pinch analysis for maximum energy recovery, ignoring cost, safety and robustness entirely.

So the stage that we are at now is that academics think that they have solved "process plant design" by splitting it in two, and then removing all of the uncertainties, ambiguities, and complexities from these two aspects.

Unfortunately the bits they removed were the important ones, and the problems they have solved were not problems at all. Optimising for any single variable whether that be approximated materials transport cost or maximum energy recovery is not smart, it is stupid. It wasn't even smart to try this, computers can only solve stupid problems.

We may nominally split the intrinsically holistic process plant design into parts, either for convenience when teaching, or for the practicalities of task allocation in professional life.

There is nothing wrong with this, but if we mistake these artificial divisions for real ones, we will be terrible process plant designers. If we do not teach students that these are artificial boundaries, they will not understand that process engineering always crosses them at every scale of consideration.

Process and hydraulic design, unit operation design and selection, plant layout, process control, instrumentation, costing, hazard analysis are all considered together and balanced against each other by professional process plant designers. They never optimize for less than three variable simultaneously. Those variables are broadly cost, safety and robustness, but these are themselves complex.

Humans however evolved to see and manage patterns in complexity. We don't need to dumb design down to the point where a computer can grind out an answer. We can intuit an answer and then apply maths and science to testing its plausibility.

Engineering is a creative, intuitive, imaginative activity. Maths and science are just two of its many tools. Computer programs are at best not quite as smart  as the people who wrote them, and long before the point where they are smart as people, become too complex for people to really understand.

The oversimplifed modelling based approach to chemical engineering espoused by many academics is a dead end, missing the point of the exercise entirely.

Sunday, 19 April 2015

Expert Witness: You Pay Peanuts, You Get Monkeys

Cartoon: the judge didn't disqualify the expert..... 

We have had a bit of a rash recently of "expert witness" enquiries which sound promising at first, and them we find out that the enquirer wants me to basically sign off someone else's plant design for £1,500.

The enquirer has usually sent this out to eight other people, none of whom has an engineering degree. We never win at this kind of thing, so we decline to quote nowadays. Chemical Engineers are expensive, and monkeys are cheap. We can't hope to compete on price if quality isn't considered.

The reason I am being asked to sign off the design is often because the original "designer" doesn't have a relevant engineering degree, and some regulatory body or insurer is feeling a little nervous.
Speaking of feeling a little nervous, it often turns out that the plant in question has actually been built, and doesn't work.

Only a fool would guarantee the validity of a design which has been built, and has failed to meet performance specification, but you will find people willing to do exactly that. You will find people who say they are willing to go to court and swear to that.

In fairness to lawyers, it can be confusing that the people in question seem to have some kind of qualification, some kind of PI insurance, and some kind of experience. It can also be confusing that engineers work to a different standard of proof than that used in court.

Let's take the last of these first. Many prospective clients, both for expert witness and troubleshooting assignments apply the kind of logic to the issue of whether a plant works or not which is applied to everyday life. When I ask if the plant works, they tell me it mostly works, or most of it works, or it works except for those few occasions when it doesn't. Alternatively they tell me they never tested it, so they assume it does.

This is however not what an engineer means when he asks if something works. What he means is has it been proven to meet the specification. This usually means has it been shown to produce the specified product 24/7, without fail.

If you haven't tested it enough to offer a statistically significant answer, the correct answer to the question "does it work" is "I don't know " if you haven't tested, (or if you have tested and every sample passed, but you didn't have a proper sampling protocol), or "No" if you tested, and a single sample failed.

If you have tested at random intervals and enough times to obtain a statistically significant answer (based on analysis by a NAMAS or similar accredited lab), the correct answer to the question "does it work" is "Possibly/Probably" if every sample passed, depending on strength of statistical significance. (Which probably in my view matches the UK court standard of proof for a yes)), "No" if a single sample failed. (The picture is slightly more complex if there is an allowable percentage of failures as with water PLCs).

Whatever the basis of the specification, every single client I have ever had an enquiry from is actually in a "No" situation, though the majority are to some extent "in denial" about this. What some of these clients want is someone who looks good on paper to write a short letter offering a justification for the client's misguided view. As such a letter is worthless, £1,500 is a pretty good price for it, as long as you have no professional ethics, professional indemnity insurance, or professional reputation to consider.

Taking these in turn - I am a Chartered Engineer, and a Fellow of the Institution of Chemical Engineers, a "Chemical Engineer of Distinction" as is says in the IChemE's description of the grade. I adhere to the IChemE's code of conduct and professional ethics.

In the UK, anyone can call themselves an engineer. British Water run a two-day course which allows people with no other qualifications or experience to officially describe themselves a "Qualified Service Engineer". I know that some of those invited to bid for the kind of work described here have this as their sole "engineering qualification".

There is nothing wrong with the qualification when used as intended to control the quality of those who maintain basic domestic package plants, but getting FIChemE required four years at university, and twenty years of post graduate experience designing and troubleshooting full scale plants.

I could buy general "Environmental Consultant" PI insurance with a nominal claim value of millions of pounds for less than £100 per year from companies on the internet I have never heard of. Instead I buy policies with a maximum claim value of a few hundred of thousands which cost me thirty times as much (as we continue to cover all previous possible liabilities, as well as all potential future ones) with companies which are still likely to be in business at the point where potential claims might arise.

The thing which really costs me money is the historical cover, and insurance against the possibility of pollution incidents arising as a result of my advice (a thing which has incidentally never happened). To compare like with like, I would advise lawyers to check for these provisions. Not having them keeps costs down.

As I have a professional reputation to consider, I cannot afford to play the numbers game that the monkeys do, assuming that as 98% of cases never actually involve expert testimony, they will not be humiliated in court by a barrister supported by a real expert. They can afford to write partisan reports which would be laughed out of court, and take a small chance of a public dressing down from a judge. They can also take a chance that someone will attempt to claim on their mickey-mouse internet PI policy.

I am very interested in real expert witness work, being asked to do the proper work of an expert witness, offering impartial and genuinely expert advice to the court. I understand the difference between fact and opinion, and consensus and fringe professional opinion. I understand the various standards of proof of the Courts of England and Wales.

I understand all of this because I have spent more time and money on courses in how to be an expert witness that the monkeys have spent on their "engineering qualifications" and PI insurance combined. I have taken courses in Part 35 compliant report writing, conducting meeting of experts, and appearing in court. I have also successfully practiced all of these things.

You can find someone to agree with your favoured interpretation for pennies, but when the wheels fall off your case in court, and you find that their PI insurance is as worthless as their engineering qualifications, you might wish you spent a little more and got expert advice from a real engineer.

Friday, 17 April 2015

What's New in Chemical Engineering?

I have finished my first couple of passes at Mecklenburgh's Process Plant Layout, and I have had to be strict with myself with respect to my inclination to leave in original text and offer comment on it in certain areas where there has been extensive change over the last thirty years, and others where there has been none. People will however not be buying the book for a history lesson, they just want to know how to lay plant out, so I will get this off my chest here.

I noted in my last post how much more significant HSE legislation had been as a driver of change in professional design practice that advances in computing, but now that I have the whole book straight in my head, I can see a bigger picture.

The most notable thing is the lack of change. Mecklenburgh was a bit of a futurist, and I have had to cut out a lot of his speculation about how things were going to change, mostly because his expected changes have not taken place. I still however commonly hear my academic colleagues telling me that these very same changes are just on the horizon, as they were in 1985.

Mecklenburgh's focus on plant layout seems to have been instrumental in him not imagining the largest change which did happen, which might be thought catastrophic from his point of view. He taught an entire module on Plant Layout at Nottingham, and this was common practice for years afterwards. I am not however aware of any UK university which now teaches a plant layout module. Research into and teaching of Plant Layout has not just changed over the last thirty years. It has more or less entirely disappeared, even though it is just as essential to practising engineers as ever.

Technology other that microelectronics has changed relatively little. Vacuum drum filters with filter aid don't see a lot of new installations nowadays, and the membrane filters which he does not mention at all are a lot more common, but almost all of the kit in the book is still the bread and butter of process plant design. We still design and lay it out in the same way.

It is amazing to me that the desktop computers of Mecklenburgh's time which must have had 6 Bit Intel 80286 processors could do all of the things which the latest 64 Bit 80663 processor can. The modern chip is just faster and has a larger memory - it's no smarter. None of the things which Mecklenburgh forecast computers would be able to do in future in the field of layout have happened, and the things they can do are are constrained by data entry requirements as ever.

Microelectronics have however done two things which Mecklenburgh did not see coming- broadly networking and PLC control, and the implications of these things.

There is no mention in the book of a software engineering discipline in 1985. Starters and controllers were field mounted, and the instrumentation and electrical "departments" were responsible for designing and programming control loops. Control panels were dumb monitoring stations.

A number of things have changed in the professional world since then (though many of these are not reflected in academic syllabuses) Motor control centres now contain smart starters and instruments, each having far more computing power than Mecklenburgh's desktop 286 PC, or even his room-filling "mainframe". These smart starters are controlled by even smarter industrial computers called PLCs.

There are very few design offices in client companies any more, for the reasons I will cover in the next section, let alone sub-departments with process, electrical and instrument design capability.

Far more profound has been the impact of networking in general, and the internet in particular. The most immediate impact has been that we can have a design office in India if we like, which we can contact as readily as one on the other side of the site, and pay staff Indian wages. Drawings can be sent back and forth, shared in multiple copies, edited and marked up in electronic format.

The second order effect of this networking has been a reduction of the cost of entering markets  which has meant that the only monolithic vertically and horizontally integrated companies nowadays are those trading in network goods. Ford doesn't make its own steel any more, and BP does not design its own plants, but Google and Facebook are huge. There are no departments any more, as there were in Mecklenburgh's day. Engineering companies specialize in their core business nowadays.

So structural changes in where engineering happens which originate in technology away from the discipline itself have been far more significant to professional design practice than any new technology within the discipline.

Though they love to try, academics are generally terrible at forecasting the future, as they are specialists. They can be relied upon to forecast that their narrow specialism will become the basis of the future of the discipline, if not the whole of human society. They never seem to imagine that their specialism will become irrelevant to practice or extinct in academia.

Academia is a sideline for me, so I'm not going to forecast how things will change. I have however noticed that engineering has changed far less than expected, and where it has, in ways which were not envisaged in Mecklenburgh's book.

The changes which have happened were driven by wider social forces, and this is how it should be. Engineers serve society - we don't tell society what to want as much as we make its dreams come true.

Similarly, it is the job of an engineering education to teach students how engineering is done. Academics will need to get a lot better at forecasting the future before they are in a position to teach what engineering is to become. As it has changed so little for the last thirty years, (and for many decades before that) our first approximation must be that the practice of the future will be in essence very like today's.

What is particularly sad is that so many in Chemical Engineering education are still claiming that the near future will match the predictions of Mecklenburgh's time, with computers doing engineers' jobs rather than just helping.

It is also commonplace in academia to ignore the changes which have happened since then, (as they do when they pretend that there is still no such thing as a software engineer, so chemical engineers will need to write their own control algorithms).

Saturday, 11 April 2015

The Next Book: Mecklenburgh's Process Plant Layout Second Edition

My next book will be a update of Mecklenburgh's Process Plant Layout, and I'm more than halfway through my first pass at this now.

Even though the book was written by a committee, in a hard to read style, and based on 1970s technology, legislation and professional practice, 95% of it is basically solid.

Fixing the writing style was trivial, so on one level the book is more than 50% written before I even have my advance.The remaining 5% is however going to take up 95% of my effort. It's just like the plant design process-the first pass discloses the genuine problems.

It is surprising just how much more significant changes driven by health, safety and environmental legislation have been to professional practice over the last thirty years than the use of computers in general and computer modelling in particular.

Mecklenburgh's book was expecting exactly the same changes to professional practice as a result of computer modelling as people are now. He was claiming exactly the same functionality as offered by the latest modelling programmes was available back in 1985. (It just took a week to run the programme back then)

Modelling and simulation have however not transformed professional plant design practice as he expected. This idea appears to be just as much the future of the past as The Jetsons.

What has transformed practice since 1985 is legislation driven by society's wish for a safer, greener world. Turns out we didn't as a society want flying cars half as much as we wanted our kids to be safe.