What is typical Plant Availability for Crude Oil Storage Tank Terminal

[rusman]

Hello my virtual friends and expertise!

We are currently performing feasibility study for the crude oil storage tank terminal and pipeline. In the terminal there will be few storage tanks (more than 30) and a series of pumping station. The required no of pump is series is 13 nos (assuming all pumps are running at the same time without no sparing being adopted YET). The pumps are strictly designed as per API 610 latest ed (12th Ed).

My questions:

1. What's the typical Plant Availability for typical Crude Oil Storage Tank Terminal worldwide?
2. How much is the sparing required that I need to consider to purchase the pumps in order to provide higher availability? Does it N+1 or N+2 or N+3 or ?

Appreciate advice from all of you expert in this fields.

Rgds

 

[ashtree]

What you are asking is a very complex problem and i don't think there is any easy answer.
There are many things to consider. The following list is just a few to start with.
1) How much duty will the other pumps being doing. If its close to 100% then there is no spare capacity so you may need more than one, but if it was only 75% than it may just be possible to run the remaining pumps longer for a while until repairs are made.
2) What is the availability of spare parts , skilled labor or perhaps replacement pumps.
3) What are the consequences of having to operate with less than 13 pumps on line.
4) What proportion of the total project cost are the pumps. If it was small than you would probably go for at least N+2.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[LittleInch]

So are there 13 pumps in series on your pipeline or 13 in side the terminal?

Typical availability is 95 to 98%.

98% is 7 days a year out of operation.

you need to find the meant time for failure rate and do a RAM analylsis if you want to get fancy.

what happens if you lose one pump? total shutdown on reduction in throughput?

N+1 is very common, but 2 x 50% or 3 x 33% is also considered on the basis that loss of one unit doesn't stop everything, mjust reduces throughtput until you can fix it.

How often do you need 100% of flow?

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

 

[pennpiper]

rusman,
I think your situation may turn out to be more complex than you think.
You say you will have 30 Storage Tanks but you did not say how big these tanks will be.
Size can make a difference.
How big will the tanks be?
Will each Tank be in it's own Diked area?
What Crude will stored in each of these Tanks?
Will it always be the same Crude or will there be a different Crude from time to time?
Will an incoming Crude to a Tank require a settling time (water separation) and Testing?
Will there be a need to "Blend" two or more Crudes?
Will there be different deliveries being made at the same time?
Will there be different Pipelines?

It is possible you will need two Pumps for each pair of Tanks/Crude Products.
It is also possible you will need a very complex Manifold system so you can be totally flexible.
Think about it

Sometimes its possible to do all the right things and still get bad results

 

[mk3223]

From the investment point of view, the factors to have the spares could be the initial budget limit, the investment return, the loss profit opportunity (LPO) cost, etc.
If feasible, a common spare pump for the multiple tanks can be considered with interconnecting piping system to have the operation flexibilities.

 

[rusman]

Thanks all for the response.

Ok, we are doing the feasibility study for the crude oil terminal and pumping system with pipeline across the country.

I know in order to get the fairly accurate data for the PA (Plant Availability) is to perform RAM study, but we are at the early feasibility study. We need to know the estimated PA based on the industrial practice and based on current operating terminal around the globe, if one could have access to it and willing to share such info.

To answer such question, let me brief about the project briefly;

1. We will need to pump 2 MMBPD of crude, with 30 days of storage. As such we need to have all 14 pumps (at one of the pumping stations, for an example) to e running all the time in order to meet 2 MMBPD of pumping flowrate. Certain pumping stations only need 7 pumps in parallel due to different in pumping head. (Pressure profile was calculated based on different pipeline elevation and length).
2. Each of the storage tank having capacity of 0.5MMbbl of Crude Oil. So there will be 2 storage terminal (equally distributed) to store the crude for 30 days.
3. Crude blending is an option, in order to get better crude oil pricing, but it will not affect the PA that we discussed about.
4. The pumping station will be in one area, so it is not directly connected to the storage tank, as the storage tank will be above ground storage tank, and designed as per the Code, with proper protection eg dike etc etc.
5. I would agree for the typical PA of 95% as mentioned by LittleInch. But 98% is too high, well though subject to RAM study. The only component is the pump (motor driven, with power from National Grid). Each pumps is driven by approx 2 to 3 MW of VFD motor. It is huge pumps and huge facilities!

Rgds

 

[georgeverghese]

API610 compliance is all well and good - pump shaft seal configuration and seal flushing scheme selection can have a big effect on pump reliability for instance - this should be reflected in the unit availabity numbers for the pump being considered.

 

[BigInch]

Pumps are RAM details that can be ignored for the moment.

FIRST determine the maximum pipeline flow rate that you need then determine the number of active and spare pumps and valves, or whatever else you build into these terminals, so that they are consistant with the specified availability.

You Said: Thirty days of storage, which I assume you mean to be 1 x 30 contiguous day period during the year.

Typical initial study of a pipeline and marine terminal with the capacity that you mention is 90%, as was used for a similar project I did a lot of work on, BTC Pipeline 2MMBOPD from Azerbaijan to Turkey. You haven't told us if this is a marine terminal, or something else, but with that capacity I have a hard time thinking it is not a marine terminal, so I'll use 90%. As you can see, that would be 10% of 365 days/yr = 37 days, which is already pretty close to your (apparently) given required storage condition of the 30 day storage requjirement.

Next, and this is where big numerical mistakes get made, determine the pipeline's maximum flow rate (as opposed to it's nominal flowrate of 2MMBOPD). Will it be production driven? If the oil field max design production rate will atually be 2MMBOPD, the pipeline flow rate cannot be the same 2MMBOPD, because if it is not working for 30 days year, you will have a shortcoming in deliveries of 60MM BBLS by the end of the year. To be able to fill your tanks and keep on operating the terminal while the pipeline is down, you must be able to AT LEAST move (on the average day) the year's production of 2MM x 365 = 730MM BBLS /335 operating days. So, in that case the pipeline's required maximum flowrate is 2.18 MMBOPD. It is more than likely that the politics of the project have already become confused and they mean both a 2MM Production Capacity served by a 2MM BOPD Pipeline capacity is possible, but with downtime considered, you can see that it is not possible. The reality of the situation implies that either the pipeline is 2.18MM/D or the aggregate maximum design oilfield production capacity is only 90% of 2MM = 1.8MMBOPD. That 10% difference isn't usually very important, until you get to these large pipeline sizes, where it can mean another pump or two, so decide beforehand what design capacity will be based on, production capacity, or pipeline capacity. The political entities can use whatever numbers they choose for the nominal capacity as long as they don't design the system for the nominal capacity.

If your pump & power units have a combined reliability of 95%, then, and to keep it simple for now, if you have pumps with the full capacity of 2MMBOPD for each (yes that's a bit of a stretch, but we're keeping it simple for now) then just with 1 pump you already have a gross reliability of 95%, which is better than the 90% terminal-pipeline gross availability you need, as long as both are running simultaneously. Which they pretty much do, since the gross 90% availability of the system was based on around 50% of pump-pipe sytem working (at 95% availability) and 50% terminal working (also at 95% reliability). Should both be needed to be working at the same time, then the probability of that happening is 0.95 * 0.95 = 0.9025 = apx. 90%, which is how that initial gross availibility you started with to roughly get 30days storage came about.

Now the equipment details

Now you need to keep the pump/power unit reliability to at least 95% in order to stay above the required 90% pump-pipeline-terminal gross system reliability.

1 pump running has a 90% chance, but would have a chance of being down for 10% of the time.
Probability of 1 pump not running is 10%
2 pumps of full capacity
Probability of 2 Pumps not running at same time: 0.1*0.1 = 0.01
Probability of 2 pumps running at the same time = 1 - 0.01 = 0.90
So that says to keep >= 90% reliabililty for pumps you need at least 2 full capacity pumps.
but we need 95% reliability for the pipeline and pumping component (pipeline is essentially 99.99% by itself)
3 pumps Reliability = 1 - (0.1*0.1*0.1) = 1-0.001 = 0.999
99.9% reliability is greater than 95% needed, so if we had full capacity pumps, we'd only need 3. Of course the capacity is too great, so I'll let you do the math for smaller capacity units.

Do the same to find the number of tanks needed, assuming you need some time for tank cleaning, or whatever. 60MM BBLS tank capacity is needed, assuming you will do no tank cleaning.

Meters?

If you have power reliability problems, do a driver reliability study to see if you need diesel generators and how many of them and electric pump drivers , or if you should use diesel fueled pump drivers, etc.

Still with me?







"He's declaring war on the planet itself."- Vicente Fox

 

[EdStainless]

I would also ask what the financial penalties are for not being able to deliver at the required time.
It may be that with one extra pump you can deliver 90% of flow at 99% reliability, or 100% of flow at 96%.
See if that difference costs them anything.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube

 

[BigInch]

Isn't that the same thing?
"90% of flow at 99% reliability". The method I outlined above calculates flow delivery capacity of a group of pumps based on the reliability of each individual pump. If group reliability is 97.5%,l then flow delivery capacity is 97.5% of pipeline flowrate. To make financial decisions, we could multiply probable capacity by price_per_unit_volume_delivered, including any penalties for failure to deliver a certain contracted volume if any. We could get the probabilities of reaching various income streams.

Otherwise, what would the difference be, or how would the result be different using the method you describe? What's the math look like there?

Richard Feynman's Problem Solving Algorithm
1. Write down the problem.
2. Think very hard.
3. Write down the answer.

 

[BigInch]

With valuable products, you should always have lots of pumps and spare parts. If you owned the oil being transported and the pipeline, just one day operating at 50% capacity on a 2MMBOPD pipeline would today cost you some $50,000,000 of your gross revenue, so having a reasonable number of spares available is obviously good business practice.

Richard Feynman's Problem Solving Algorithm
1. Write down the problem.
2. Think very hard.
3. Write down the answer.