UKH

Out of curiosity and more from a maths perspective than a project fear exercise, I was wondering about the knock on effect of port delays, and how you would actually calculate them.

Starting with 1 lorry being delayed for an hour starting at day 1 hour 1, and 100 lorries an hour (as an example) leaving the port, would the knock on delays/queues rise exponentially? (my maths isn't that great)

Would we eventually run out of lorries?  and would we need to keep adding more into the system to keep the supply chain from slowing.  

My poor data source is from the best scenario I could find (with some saying much longer to process). I seem to recall a driver the other day saying over 100 left dover each hour.  My calculations based on Dovers figures divided by hours in the year had it about 300, so 100 should do for now.

I know the UKC collective has some good mathematicians, so I though why not give it a go.

No rants about the situation, just the logistics of it please : )

How many lorries are on 1 ferry, how many ferries / day?

Then work out how long you've got per lorry to clear one ferry load before the next one arrives.

Right... just got some added info:

Every day, an average of 10,000 lorries pass through the Port of Dover and are typically processed within two minutes.

The problem is not clear, why would a delay on one lorry impact the others ? Are they in queue ?

Yes, all the lorries leave Dover port as they cant handle the extra processing and will be stacked up on the M20 (I think thats the motorway) to be checked and let go. With plans to use Manston Airfield.  

So you have: (1440 minutes in a day) x (the average number of inspection channels open at once) minutes per day, to clear lorries assuming 24 hours of operation. That implies that at the moment about 14 inspection channels is enough to clear all the lorries.

If the time to get through inspection rises to 60 minutes you need to multiply that number of inspection channels by 60/2=30 to ensure you can still clear the lorries fast enough. Otherwise as time goes on a linearly growing queue of lorries will build up. This queue will continue to increase in size linearly with time until the number coming in falls or the rate of inspections increases.

Prizes for working out which of those two outcomes (lorries from the continent stopping bothering to get on the ferries to come here or the home office managing a 3000% increase in inspection capacity) will happen. (apologies for possibly violating your rant rule...)

The branch of mathematics you are looking for is called queueing theory:

https://en.wikipedia.org/wiki/Queueing_theory

In a simplistic case if every ferry arriving and leaving after a certain moment is delayed by a fixed time (say 1 hour) then we have effectively lost the number of ferries that would have arrived and departed in that time (say 1 each way).

Therefore on the UK side we would have 1 ferry load fewer inbound lorries coming through customs, resulting in reduced queue times. We would have 1 ferry load more outbound, so initially the queue would increase by that number. Assuming the ferries are not at 100% capacity all the time the queue would reduce every time there was extra ferry capacity until it reached a steady state the same as before.

If process times (e.g. at passport control) go up then the carrying capacity of those processes go down (unless extra staff are laid on). If demand is above capacity then the queues will grow indefinitely.

Generally the average length of a queue depends on the "lumpiness" of items (lorries) arriving at the back and being removed from the front.

Edit: crossed posts with kestrelspl

Yes, I'm guessing that the Motorway wont cut it.  That the motorway will be the stack for transfers onto the hardstanding at Manston where a channel system could be introduced.

I believe they have been trying both free running of the lorries in and out, and a convoy system.  

So another spanner in the works if the no. of channels is unknown.

This answer is a lot more accurate than mine. I mistakenly read your average queue time for an average time for the inspection process to take place so it's harder to give exact numbers. The rough conclusion remains the same though, the queue will continue to grow (hard to tell the functional behaviour of the growth without more queueing theory knowledge) until the rate of inspections increases or the rate of lorries entering the queue decreases.

you have to account for the mean time between failure of stacked lorries in lthe parking lot as with that number of lorries the vehicles and personnel needed to carry out repairs, deal with illnesses or driver problems may disrupt the process.  Add any problems from inclement spring weather, storms etc and also the effect of any bottlenecks that arise due to the buildup of vehicles.

Also dealing with any opportunistic stowaways that use the disruption to conceal themselves.  Overheads could escalate.   Perhaps even affecting the supply of spares needed to deal with breakdowns.

simples

I hated maths at school, but if it was problem solving like this I'd have been happy - rather than calculus and algebra.

I recall reading about the maths behind motorway traffic and the variable speed limits, which was interesting.

Excellent - thanks.

Look up Fermi problems, should be right up your street. e.g. How many piano tuners are needed in Chicago:

First work out relevant variables:

1) How many pianos? We can get that from (population of Chicago/people per house)*fraction of houses with pianos

2) How often does a piano need tuning?

3) How long does it take to tune a piano and then get to the next job?

Multiply the three answers together to get the number of hours of gainful employment for a piano tuner. Divide by 40 to get the number of full time workers you can support  

I will also add that by doing this you're essentially just doing algebra but where you've substituted the word "x" for a variable you've described in words. Calculus is also tremendously useful for solving problems where one of the variables you happen to need turns out to be a rate.

Order of magnitude calc :

10,000 lorries parked up is 100 miles long.

The M20 'car park' is 50 miles long

It's another 25 miles from the end of the M20 to Westminster

I reckon they have about 18 hrs before the 'logistics issues' are parked outside their door. 

Just found out that the Manston stacking is for outgoing lorries.   Imports will be on M20.  3 lanes... officials swarming like flies!   Probably will have to have bays, but that relies on all lorries taking equal times, which is going to vary.  Its going to be like the post office queues - yours always goes slower.

So still calculating the variables. 

Maybe I just had a rubbish teacher who couldn't make it interesting!  I'll look up the Fermi problems.  Good for brain exercises.

Maybe a good one to just simulate in Matlab.

Using the figures, if 14 lorries were processed at a time, and the amount of lorries leaving the port was 160 per hour to the M20 stack, the net build up would be 146 lorries each hour, equating to a backlog of 3,504 vehicles in 24 hours, which given the length of an arctic and trailer stands at about 12m, takes up about 2/3rds of the M20 stack in the first day.

So given the figure of 60 lorries needing to be processed simultaneously to keep things moving, the questions that should be asked in the House are how many staff are being trained and how many will be there on day 1 perhaps.

I'm guessing that the outward journeys won't be that bad as the UK government want the trucks to hit the continent as quickly and hassle free as possible.

The problem will be at the other side. Will countries such as France be as receptive to wanting to clear trucks as fast as possible to enter the UK once we crash out?

And would the French be as strict with stopping illegals entering the UK?

If ferries are delayed at France would that affect your initial calculations?

Possibly so, as its all part of the loop, although more research shows that Manston will be a checking point for exports and the lorries will be released in batches from there to the ferry port.  But I would guess new procedures at Calais too.

I can't find any data about HGV spare capacity to do further stats.

I guess it depends if Let Touquet accord stays in place. At the moment French do French customs in Dover and we do British customs in Calais. Changing that arrangement presumably requires actual construction work to put check stations in where people come off the ferry or very convoluted lorry staging somewhere that's not the port whilst still guaranteeing you can't accidentally get into the UK without checks by taking a wrong turn.

For my sins I did a course on queuing theory as part of my Computer Science degree ages ago.  What I remember is that when you start modelling the arrival rates as statistical processes you get into nasty maths really fast.   The other thing I remember is it isn't nice and linear, you get exponentials: if you want to keep the servers busy your queue length gets really long really fast.  If you want to keep the queue length short you need a lot more servers than you think.   Fairly quickly to solve these things in real world situations you need to move from trying to do algebra to running simulations.

In the Dover scenario there are complicating factors beyond classical queuing theory:

a. The long queues will eventually hit junctions with other roads and interfere with traffic not going to Dover.  The road system will probably gridlock.

b. If there are significant delays then each lorry moves less goods per week so you need more lorries to maintain the same throughput.   Somebody needs to buy all those lorries and hire the people to drive them and the extra lorries themselves will use road space.

c. There are network effects in the supply chain.   If you are making something by assembling ten different components and one of those components is hugely delayed then you can't make your product.    A factory could be getting 90% of the supplies it uses from within the UK or through other ports but if one essential component comes through Dover it would still be screwed.

That is going to be chaos. The road to Manston is one step up from a country lane...

Almost two steps up... It has white lines down the middle and everything 

I noticed that from the video, and busy at rush hour.  I was just thinking.... winter, 24/7 HGV.  Could the road cope before crumbling.

Yes, b.  I'm still looking for how much spare capacity the UK has before running out of available freight.    I'm not sure how much the supply chain relies on cabotage for EU freight runs.

c. From a report - Honda UK has 350 lorries delivering parts each day to keep its plants running.  They only have space for about an hours worth of parts, so the Just in time supply chain is what works for them.  Delays would certainly mean production halts/lay offs.   

The numbers certainly went a bit wild for me when I played with the extrapolations, it was like wtf!

Ignore Brexit initially.

It is delivery time that has increased. Production isn't demanding extra parts and the supply isn't drying up. If they modify their supply chain or just hold 1 day of stock. Zero problems. Bad weather can delay shipping,  power failures delay trains, chunnel fire, rail strikes, French port blockades, 2" of sleety snow  etc.. factories don't start laying people off now or in the previous 20 years.

Millions of containers arrive from outside the eu every year that have already had their customs and tariff work carried out. The eu currently applies tariffs to many goods arriving in it we are just oblivious to it until we buy something direct from the USA for example, so there isn't really a change in many respects. 

But I think we are talking short term here (but a week is a long tie in queuing theory).  What will happen while any form of procedure gets implemented. I think all the heads of industry accept that its short term not long term but in the initial period, any system will cause queues, whether it be security or paperwork.   Locked and verified shipping containers/manifests arriving by ship are a bit different from the artic truck load involved in Just in time supply chains. 

My post about Honda was direct from Honda themselves who stated they only had capacity to store 1 hours worth of parts - so I cant comment further on that.

I'm more intrigued - from a logistics point of view the effect of a small hold up at point of entry, and to why the government had been trying to offset the immediate problems by advising the storage of up to 6 months worth of stock to industry.  I'd be satisfied if I thought there was a system agreed, but all the various industry leaders who would be affected say not.

Anyways - back to the maths!

If Honda currently only have 1 hr of stock... how in the past have they coped with bad weather, strikes, fires, power failures, port blockades, congestion or even a lorry just having a puncture or breakdown? 

Have you ever played SimCity or Cities Skylines?   Watch what happens when you have a really well functioning city generating lots of tax revenue and you change the main freeway entrance from a wide highway to a minor road.

The lorries back up, all the roads near the entrance grid lock.  Police and ambulances can't get about, factories shut down, shops cant get supplies, property prices fall, people start leaving.   When the factories at the bottom of the value chain don't get the raw materials the ones higher up can't make the products that depend on components from  the first level factories. The whole thing goes from highly optimised to total sh*t with huge fixed costs and no revenue and losing money really fast.   

If you wanted to make the city work with the more limited connection to the outside world you'd burn through huge amounts of money reconfiguring your roads and getting a different balance of industries.  You'd almost certainly run out of money and need to borrow.  Then you want to have low taxes to try and stop people and businesses leaving but you are forced to raise them because you need to keep up payments on the loans and your costs are still high from all the stuff you needed for a more populous and successful city but your tax base is smaller because there are less people and industry.

Highly optimised is not the same as robust and tolerant of disruption: in many ways they are opposite goals.  This isn't something you want to try out in real life based on total hand waving assertions that it will all be fine.  It could be a f*cking disaster far worse than anybody imagines based on naive linear modelling because of the network and feedback effects in a complex system.

perhaps catastrophe theory might help

Also semiconductor effects, the transistor effect of a small current flowing up the outside lane in a traffic constriction having a pinch-off effect on the large main current in the "output circuit" formed by the bulk of the traffic in the other lanes might be an example

https://www.tandfonline.com/doi/full/10.1080/2058802X.2018.1452422

In case you are interested...

By waiting for the short term, localised problem to go away.  Which they typically do in a few days.

Here it’ll take weeks or months to restore normal flow, and it’s not happening to one factory or one company, but to them all.  For months.  

This introduces non-linear coupling of effects through (a) widespread disruption to the logistics network that all the individual queues run on and (b) financial losses far beyond the typical contingency funds and potentially (c) coupling with civil unrest.  What’re all these people going to do when their factories are shutting down?

That’s Bad.  It’s also much harder to understand or model.

Edit - ooo look a dislike!  Strong stuff.  Is the the best you’ve got?  Excuse me for being peevish but it’s hard to watch this process unfold.   Edit:  a second one.  Keep it up!  Save some of that Dunkirk spirit though, you might need it for more than just pressing a button...

So Honda copes just fine. As i said, demand and supply haven't changed, only deliver time. If on the 28th March you ordered a double delivery, you have a buffer. But i don't see it being a problem, because of the volume of controls carried out at the departure point. 

Ps. Personally I'm quite happy with any sort of customs union etc. I'm not advocating no deal/ hard Brexit as the best or only option. Just discussing the logistics. 

Why haven't all the more advanced countries around the world that aren't in eu been cast back to the stone age then...? They seem to manage? Perhaps there are just different ways to do business and trade? 

Because none of them suddenly severed ties with their major trading partners, let alone without having any sort of competent plant in place first.

Further, there are only 3 non EU countries in the whole world more advanced than the UK by GDP. We have nowhere to climb to but a long way to fall.  It’s hard to see sudden and drastic change improving things, and easy to see it disrupting them...

They are also currently working on part of the M20 to make it a smart motorway (provide an additional lane I assume), having driven past yesterday that certainly wont be finished before end of March.....further chaos to add to the mix.

Look up the Dunning-Kruger effect you might find it illuminating.

Obviously you can build up a city with only a narrow connection to the rest of the world.  You can set it up to be self-sufficient and you can set up the road network within it to work well in that context.   It might not be as successful as one with a lot of trade to its neighbours but it can do OK.

That's not the point.  The point is that if you start from a city where you have spent a lot of time getting everything set up to work really well when there is a high-bandwidth connection to its neighbours and then you take that connection away or make it much smaller than your city will collapse.

You can try and recover by changing everything around within the city to be more like the one which doesn't rely on a high bandwidth connection but it is not easy.  It costs huge amounts of money to reconfigure,  your tax revenue is a fraction of what it was but your fixed costs are the same.   You try slashing expense on health and police and services like garbage because you can't raise taxes or people and businesses will leave any faster.  You take out loans to pay for the new infrastructure and much of your taxes go to paying them back.  Usually you go broke.

Nothing new there. It's just another flavour of;

Unconscious incompetence 

Conscious incompetence

Conscious competence

Sub or unconscious competence

Or not. BMW are shutting their Mini plant in the UK for at least 4 weeks AND expanding their Mini plant in Holland (and diverting new production plant to Holland). Bet that gives the guys in Cowley a warm and fuzzy feeling.

But the roads don't change. If you have a 100hr journey and you have a 10hr delay, it's a factor that can be managed through change X and y to take up or increase by 10%.

I'm not suggesting it's a good outcome, but millions of ton of goods arrive monthly from outside the eu every month. There are processes already in place. I don't personally think it will be such a massive problem. It's just my view.  

How do we not know that lots of companies see Brexit as a perfect chance to restructure and blame somebody else? Neither of us know. 

You mean like BMW planned to build the electric version of the Mini in Cowley but have signed up with Geely in China to develop an electric vehicle range?

Companies don't need excuses to re-organise, they just do it.

I'm not even sure if this can be solved with the information available. Each lorry represents a number of purchase orders which will be fulfilled when the goods are delivered. What we don't know is whether or not a delay in fulfilling an order results in future orders being brought forwards or postponed.

I think to answer your questions. The delays wouldn't rise exponentially. We would hit real world limits (like the number of lorries available as you know) and the rate of supply would settle at a different rate. Calculating that rate though, is something that is beyond me.

What about replacement parts for the lorries?  Do they come in by lorry?

Its lorries all the way down I tell you!

 You need to simulate these things to understand that it isn't a linear system there are feedback effects.  The roads have to change to deal with the new patterns of traffic flow.

Suppose there are 10 hour delays and one lane of the M20 becomes a parking lot to cope with the queue.  Then the capacity of the M20 to move traffic is halved because only one lane is in operation - that has knock on effects.   There's no option to overtake if one lane is a parking lot and the other is choked with trucks.  If the journey time is increased by 10 hours then they probably need more drivers to keep within working times.  They need hotels to sleep in, toilets, food.   They need more trucks because each truck is doing fewer complete trips per week.   Other services get affected by the congestion including police/ambulance/fire/refuse/local deliveries/people getting to work.   Every business needs more vehicles to deliver the same service because their vehicles get less done per day.

To get the country straightened out to work efficiently in the post-EU state you will need to reconfigure the road and port infrastructure to deal with the new traffic flows and restructure industry within the UK to deal with a new set of challenges and opportunities.  It will be tremendously expensive and disruptive for a long time.  Worse still there is the opportunity cost of dicking about solving a problem that you created yourself rather than putting your time and money into the next generation of technology.

Did you read the paper I referenced? In the main it was positive  but the last part was somewhat downbeat. Ie. Even short delays in some areas means that capacity is outstripped and queues grow quickly. That capacity can only grow so much. 

Its an efficiency / resilience thing. 

Very efficient systems are not very resilient it would appear.

Dover of course will be looking at this  but there are limits and they don't always want to be upfront... They are a business too... 

You completely blew any logistics credibility I had previously given you when you said truckers need hotels. 

Thats the changes as part of operation Brock - which is what the stack system is going to be called.  I think they are making a permanent barriered off section for the lorries to stand in.

In 2019 Operation Stack is due to be superseded by Operation Brock. Work began in May 2018 on this scheme managed by Highways England, designed as a temporary solution to manage lorry queueing and traffic flow at the Port of Dover after Brexit.

In September 2018, Highways England revealed in a freedom of information request that "Operation Brock, the code name for the management of freight in a no-deal scenario, would not be automatic and would require steel barriers to make a planned contraflow system on the M20 safe for ordinary vehicles" and that "£30m has been allocated to cover the design, build and initial operation of the scheme for up to six months."

Or toilets...  

What I meant was that you might want to consider why you (someone who I'm presuming from the discussion has very little experience of car manufacturing logistics) are convinced it's fine, when literally all of the car manufacturers in the UK are saying its a massive issue. Maybe, to use your words, you're unconsciously incompetent on this matter and should trust the people who do this for a living.

The diligence of French customs may well depend upon whether or not we refuse to pay them £39 billion.

Does anyone else remember the time when the French had a dispute with someone like Japan and made it compulsory for all their import/export paperwork for electronics goods to be produced in French and to go through one particular customs office in the arse end of France staffed by one bloke and his pet spaniel?

YES! WELL DONE!

THAT'S THE WHOLE POINT. EUROPEAN COMPANIES WILL RESTRUCTURE TO ELIMINATE THE DELAY FROM THEIR SUPPLY CHAIN.

PLEASE DON'T SHOUT.

It hurts my ears.

To clarify that, it badly affected British produced electronics (VCRs etc) being made by in the UK by Japanese companies.

It's a discussion, are people who are only experts in a given field allowed to have an opinion? 

In any discussion, if person x has their view shown to potentially be incorrect or different to person y, that's the whole point of a debate or chat. It's not to turn up and just call everyone idiots or unknowledgeable (however politely you phrase it). Better to let the argument win the day, not petty name calling or belittling. 

No, never worked in a car plant. But have in the past crossed plenty borders, eu and non eu, in a car, van and truck carrying goods. So I have a rough idea of the processes involved. 

The threads here would be barren if only experts commented. 

Took a while but I've been out. Makes interesting reading.  The last part about the 2 mins extra causing delays, is probably why early on in the discussion, it was decided that the inspections were moved to Operation Stack, to keep Dover moving.

I'm not disputing that, absolutely people with all experience and none can participate. We're all guilty of being unconsciously incompetent when approaching areas outside our own expertise. I was merely suggesting that stating: "It is delivery time that has increased. Production isn't demanding extra parts and the supply isn't drying up. If they modify their supply chain or just hold 1 day of stock. Zero problems." when everyone who is an expert in the field disagrees with you indicates that you might improve your standard of debate by checking things a bit more before saying them.

For instance that statement I've quoted ignores the staggering amount of warehouses that would be needed to pop up in a short space of time, the large amount more lorries needed so they can spend that time sitting in a queue and also reconfiguration of the way making changes to product lines and assuring known faulty parts don't sit in the aforementioned warehouses is handled. All of those issues increase costs dramatically and make the UK a less attractive place to put your factory.

This.

Indeed so. There would be very few Brexit supporters contributing.

Martin

I've been trying to picture what would happen to a "roadport" that routinely had lots of lorries standing around, of all kinds of cargo in no particular order in comparison with our reasonably rapid entry system within the EU if one of the inevitable agricultural biosecurity events (of which there have been a number during my lifetime) was to occur.   eg Foot and Mouth, Swine Vesicular disease etc etc.

Given the total chaos we had with realative free movement and no heap of stacked and stranded transports occupying acres of space, what would entail if this occurred when this changes to the envisaged scenario?

How would normal supplies get through the chaos?  We are creating an enormous national vulnerability comparable.  We'd need a Britain Airlift and humanitarian aid at Dover, 70 years on.

A rather good scheme for a disaster planning wargame.  FMD flourishes more in the cold months of which March can still be one

I think if the delay is fixed at 10 hours then that wouldn't be a problem. The problem is that if the rate at which lorries can get through the port is less than the number of lorries trying to get through then the delay time will increase indefinitely.