Road tanker loading utilisation rates

[LittleInch]

I don't know if this is realty the right place, but couldn't really find any other forum although maybe chemical plant design and operations?

Anyway, I'm trying to see if there is any recognised Rule Of Thumb or experience for percent utilisation of a multi bay road loading system (4 bays).
So say each tanker takes an hour from drive in to drive our - 16 hours operation, 100% is 16 tankers per bay. But doing this for multiple bays day in day out 7 days/week seems unfeasible or very inefficient, i.e. more tanker demand than capacity and hence queues of tankers. I don't have much space for tanker queuing / parking.

This is for a distribution depot, so varied turnaround times for a tanker fleet.
It's in the UK, but should be the same for the US and Canada.
I can only imagine that this is a logistical nightmare to program in, but thankfully all I'm trying to do is size the system appropriately and looking at overall flow numbers into the storage depot.

My current thinking is that on a daily basis of two shifts (16 hours) we would be looking at 70-75% without having dozens of tankers parked waiting for space on the bays. This is for a single bottom loading product.

Any useful links or places to look at gratefully appreciated.

Thanks, LI.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

You already know there is no design guide for this, right?

I think you might be using reverse logic. How many tankers you need depends on instaneous customer demand and the average voyage time, which the max is pegged by 100% loading capacity, ie 16h op/day x 4 bays 64loading-h / 1hr/load = 64 loads in a 16h-day.

If average fill time is 1h, the avg voyage time is 2h, then Total delivery time is eh
You can send 60 loads per day. 16h - 1h average fill time 1h (you are not sending anything during the first hour) = 15h
15h x 4 bays = 60 loads.
1h filling / 3h total delivery time (count loading time as a part of total delivery time) = 0.33
4 bays/0.33 = 12 trucks
So you need 12 trucks to get 60 loads out per day.

That's everything working 100% at full terminal delivery capacity.
Lots of terminals do that 24/7
All trucks filling in 1h
All trucks round trip travel of 2h
If you want to target less then 100%, you can if you want to, but it's not necessary, unless you say, work 8h.

Let's say you target 75%.

Now whatever messes up that smooth operation will require more, or less trucks.
Weather, traffic jams, customers want more than 75% for this weekend bank holiday.
Or they want less this week for whatever reason.

Do you want the capability to deliver more than 75%, this week, or this afternoon, then you need more trucks. If traffic is bad today, then your avg delivery time goes up, you need more trucks. If delivery time drops today, you need less trucks. If delivery time on some routes is longer, but shorter on others, 12 trucks might be just fine, as long as customers are ordering their usual deliveries.

The number of trucks in line are constantly being affected by the variations of all kinds you might get on any one day that you didn't expect to get when you scheduled the truck deliveries.

As a matter of curiosity, whenever I looked on Google Earth at various large distribution terminals, I can always count a lot of trucks sitting in the line, or parked at the nearest vacant lot waiting to join the cue. I have counted up to 124 trucks, but yes, those were for some pretty large terminals that had 12 to 24 loading bays.

I think it boils down to how much you can delay orders a few days when things get bad, verses how many of your customers will switch suppliers and increase your lost sales when you tell them you can't make their delivery till next Wednesday.

With a small 4 Bay system, I wouldn't think you would have more than 1/2 the number of trucks required for your target operating delivery day hanging around at any given time, but it is a matter that is greatly affected by increasing customer aggrevation level, lost sales possibility, traffic and weather.

If you do some sensitivity analysis on the % fill time vs total delivery time, like I suggested above, you might be able to bracket in the number of trucks you need for min and max capacity, throw in some randomness and do a quick monte cargo simulation to see how bad it might get.

Maybe you can get a handle on the an average travel time by checking your time to destinations at various times of day for your particular delivery area on your Google maps route planner.

Do you need somebody to do that for you?

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[1503-44]

If you don't do a montecarlo analysis, you can determine how many trucks will likely be in the cue based on random arrival times of tankers, given a binomial (or other) distribution of arrival times around a given mean and likely variations, say within 1, 2 or 3 sigma, but that might be more subjective unless you had a means to determine something better than a std normal distribution, but it is used a lot, rightly or wrongly..

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Snickster]

Attached is a old Mobil specification that gives some guidance.

 

[pierreick]

Hi,
The production rate is the product of MDC (maximum daily capacity) time OEE (Overall equipment effectiveness).
In the chemical Industrie an OEE of 75% is standard, below this value Operation should check for losses and downtime.
In your case MDC = 16*1 hour= 16 trucks, loaded per bay, per day.
One way to improve the MDC is to review the process steps with associated time. Probably if you load from the top, you will save time, no back pressure.
Note: I believe your operation is not safe, there is no way to control the status of the bottom valve at the end of the operation, it may leak and could lead to catastrophe when the flex hose is disconnected.
Good luck
Pierre

 

[LittleInch]

Gents,

Thank you for the responses.

Mr 44 - This is really up to the operating company as I don't have a lot of the information required to come up with a judgement, but I suspect neither do they as this is a new location and a new operation for them. I'm looking more at the capacity of the pumping system and what it needs to do and then look at a bigger picture issue with what period a marine tanker needs to come in to fill the tanks as they then need to have sufficient ullage for the bulk cargo, but equally not run out.

Snickster - Many thanks for that - useful information, especially the average of two trips per tanker per shift. Clearly if two trips takes 7 hours then there is some dead time for the drive rand tanker before the next driver turns up - thankfully not my problem as this is logistics and scheduling for the operating company - I'm just trying to get a sense of overall throughput and how a multi bay system works in terms of pumping rates.

pierre - That 75% figure is what I was really hoping to find. FYI, this is pressurised product so I can't use top loading, but top loading is now not normal for environmental reasons and personnel protection from the vapour. Plus risk of falling off... They have special dry break connectors which seal before release.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

Worrying about the number of trucks standing in the que is not the big picture.

You can't even see the big picture without the market study. W/o that, all you have to go on is the 4 Bay system working at max delivery capacity, based on the estimated 1 hr fill time and 2x 8h shifts. Where did the 4 bays come from?

Ships are the big picture alright, but that picture should be already framed. But it is also not critical, unless you intend to exceed docking and discharge capacity. Within 0 and max Dock capacity, you can usually work out a ship delivery schedule that will supply whatever you need.

It would seem that the company's marketing department would have some information, if not an entire marketing plan written for the market they intend to serve. Is that not the first step in a business plan? Especially sales projection for today's quantities and at least the next 5yrs, with a longer estimate in the 10yr range. How did they even determine that they need 4 loading bays?

Pump system capacity: Why would it not be a max of what you can load in a working day? ie the 60 loads per day of a 4-bay system working your defined 2 shift plan? with a VFD you could do anything up to that limit.

Marine tanker schedule: why would you not use the same 60 loads per day figure for estimating the initial shipping schedule too?

Site tankage:

Tank sizing for ship delivery is critical, not the road tankers, but still not that difficult.
Most distribution terminals in the US operate with a local terminal storage capacity of 1 month when feed by pipeline. They will allow that to drop to a 2 weeks supply on hand at high demand season, or during refinery turn around season. For ship delivery, I have used 1 month plus a 1 to 2 week extra supply buffer. Most long route ships can easily meet their ETA within +/- 1week. If they were running short routes, not trans-oceanic, 1 month supply on hand should be adequate.

1 mo supply
60 loads per day x 10,000 gal/load x 30 days/month = 400,000 to 500,000 bbls.

SuezMax ??? 800,000 to 1MM bbls That will double your tank volumes.
Normal Ship schedule: 1 SuezMax every two months+/-

Will the road tanker bays have their own tanks, or feed directly from the ship receiving tanks? That brings us to the small picture and back to the factors I mentioned above, especially the fill time to total delivery time ratio. Whether that's 1+2 hrs or 1+7h, I have no idea, but somebody needs to know to find out how many trucks are needed. Then apply the variation to truck voyage time schedule factors to get your loading system tank size, or forget about all that and feed directly from the ship delivery tanks.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[LittleInch]

Mr 44,

Very well written and I might use some of these to highlight to the client what he needs to think about as I'm really not sure whether they have or not....

Its a bit complex and I don't really want to go to town in a public forum with specific information, but the shipping issue in complicated by tidal restrictions meaning only some of the month tankers can get in, plus the size of the cargo is about 70% of the total storage capacity, meaning that there isn't a lot of leeway at peak demand between not enough ullage and running out.... This is not normal as far as I can see, but it's what we're faced with as the storage is existing and were adding the marine capability together with a pipeline ( hence why I was initially involved but it's grown a few arms and legs since then....)

Surprisingly it doesn't look like a lot if this has been fully thought through, but maybe it's just not visible to me....

I'm trying to not overdesign the system which has already been built with two bays with the expansion initially determined to need 4 bays. I'm trying to see if the existing pumps with maybe adding one more can suffice or not. Plus how the overall system can work.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

You'll have to add storage capacity for the contiguous length of time that docks are not available and the ship is waiting to unload. Another probability distribution to add to the montecarlo sim?

I did a montecarlo study to get tanks volumes needed at Lamu.

BTW They are still looking for finance of the southern route. Western banks have lost interest in being part of another environmental mess up. Lots of opposition to that route. Who would have thought?

Sounds like somebody knows, but you are not getting the world. Ask to talk to their ops super.

You have my email. Feel free.


--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[pierreick]

Hi,
This document WCM (world class manufacturing) may help you on your project. All indicators are defined.

https://dergipark.org.tr/tr/download/article-file/618812

Pierre

 

[shvet]

Hope this will help

Step 3 - Predict arrival patterns based on local experience with customer demands within the delivery area:
1. Peak arrival periods: Trucks per hour, period duration (hours). (Morning queues at start of operation can represent greatest arrival frequency over a relatively short time.)
2. Seasonal variations: can be significant with some products; e.g., mogas, heating oils, fuel oils, asphalts.

Step 4 - Establish base case configuration:
...
4. For peak arrival periods in excess of 1–2 weeks, calculate the number of spots required for each product class based on a rack utilization of approximately 50%.
5. For overall yearly average loading rates, calculate the number of spots required for each product class based on a rack utilization of approximately 35%.
6. Select the highest number of spots resulting from the controlling case above rounded off to a whole number.
7. Attempt to minimize the total number of spots where possible by considering:
a. Re–examining requirements (peak volumes, periods, etc) which result in a high number of spots.
b. Combining low volume and/or fractional spots.
c. Two–stop loading for low volume product classes.
d. Dual service loading spots based on seasonal variations of different product classes.

 

[LittleInch]

Thank you gents, that sort of checklist will be very useful and I appreciate it.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.