Storage Point in RBD

[Christoph Frenkel]

Hello,

I`m creating an RAM-analysis for an gasification plant. For this I want to consider the storage capacities (in example a collection container). Because the storage capacitites influence the mean availability significant.
I tried to solve the problem with the "Throughput-Analysis" but the mean availability in the simulation doesn`t alternate...

How can I create a storage block in the RBD, which effect the trait of the upstream components?

Sincerely, Christoph F.

[Arai.M]

Could you describe in more detail how storage capacities affect the mean availability? In principle storage capacities affect the throughput itself, not the availability.
Regards,
Arai

[Christoph Frenkel]

Hello,

I will explain the effect of an storage point at one section of the gasification plant. The plant uses coal dust for the gasification.
Thus, the coal receiving unit, the mill and the gasifier are connected in series (also reliability-wise).
Between the mill and the gasifier is a coal dust bin located. The bin has a storage capacity, which is enough for running the gasifier 4 hours.
If the mill has an MDT of 3 hours, the downtime of the mill is completly compensated. Because the gasifier is feeded with coal dust of the coal dust bin.

Summarised, the coal receiving unit, the mill, the coal dust bin and the gasifier are reliability-wise connected in series. The storage capacity of the bin compensate parts of the MTTR or MDT of the upstream units. This effects the mean availability.

My question is, how can I implement this storage in an RBD?

[Arai.M]

There is probably different ways of addressing this. One way I see is to have the the mill and the coal dust bin in a standby configuration. The mill is the active block and the coal dust bin the stanby (with and active time to failure of 4 hours - this could be further refined). When the mill goes down, the dust bin hast then 4 hours till it "fails". Once this occurs, if the mill is not operating yet, the system will go down.
I hope I understood your question.
Regards,
Arai

[Christoph Frenkel]

Hello,

Thank you for your response! The standby-container is a solution for the effect of a storage point for one "active" component.
What can I do, if the active component is a subsystem? I will explain:
I have a subsystem (created as a subdiagramm) containing:
a series system of: mill feeder -> mill -> pump -> bag filter -> screw conveyor

If the subsystem fails, the storage point will operate.

It is not possible to choose the subdiagramm as the active component in a standby-configuration. What can I do?

Sincerely Christoph

[Arai.M]

Hello,
You cannot use a subsystem as an active buy you can put all those components in series as actives. Whenever any one of those fails, the standby will be activated.
Hope this helps,
Arai

[Christoph Frenkel]

Hello,

thank you for your response.
Back to the main question: Another way to illustrate a storage point with BlockSim is not possible? :confused:

Christoph F.

[Arai.M]

You mentioned that your subsystem is made out of components in series, is there another complication as to why you wouldn't want to put those blocks in series in the container instead of in a subdiagram? :confused:
Arai

[Christoph Frenkel]

Hello Ari!

The example with the series-system was only an example! My system (gasification plant) consist of approximately 1000 components.
In the gasification plant are some areas, which have several traces. The traces have the same function. They are splitted at the beginning and connected at the end again. (The reason of several traces is only the flow rate, one mill can`t handle the complete flow).
For example the coal preparation. It consists of appr. 50 components. They are sometimes connected in series, in parallel or in standby. For the whole area, I have a storage. The storage has a backup/capacity for 3 hours. If an component fails in the area, the storage can compensate this outage. But only, if the outage time is smaller than the backup time of the storage.
In consequence the standby is not enough! What can I do?

I hope, you understand what I mean. :o

p.s. Maybe you know the EPRI-Report AP-5276 "Availability Analysis of an IGCC"? They work with storage by calculating a larger MTBF-value.

[Arai.M]

Hello Christoph,
the principle of adding the standby is a similar concept that adjusting the MTTF. However as you have found out, this is not possible with complex diagrams. One way, like you have mentioned is to find that equivalent MTTF that will allow you to model the additional life in the system.
The other one is to analyze this exclusively from the throughput perspective. The system stops once the backlog is consumed in your case (limited backlog) however, the tp reaches zero.
Another way and again this is an approximation, is to pick key components (those that are most likely to bring the system down) and put them in the same group as an active block in a standby (with the standby again having that extra 3 hours of life). If any of those critical blocks fails, they will trigger a pm that will bring the active down.
Ideally, you should be able to add a subdiagram in a container which would allow you to do this without approximations but currently this is not available in BlockSim.
Regards,
Arai