Operating Hours for Plant Capacity 3

[sshep]

My Friends,

We were issued a plant design from the licensor which is based on 8000 hrs/yr. This low operating hrs has raised many questions because it implies a stream factor well below what is expected for this type plant, although the total annual capacity is exactly what is needed.

Is there any industry standard for choosing operating hours per year for name plate capacity? Also what reference would give definitions for availability, reliability, and related stuff?

Any help is appreciated.

best wishes,
SShep

 

[Latexman]

It varies by continuous, batch, process, technology, etc. In my industry 300 days/year is conservative. It is batch process, heavy fouling, lots of cleaning.

For continuous process the "textbook" answer IIRC is 8000 hr/yr.

For world-scale mega-plants like refineries they plan on a 2-4 week shutdown every 5 years type thing, so 8760 for 4 years and 8086 on the 5th year.

Good luck,
Latexman

Technically, the glass is always full - 1/2 air and 1/2 water.

 

[bimr]

If you are operating less hours, would not the stream flows be higher for the same overall capacity.

As latexman posts, it is difficult to come up with the typical working hours when the production process varies so much.

The information may exist in the trade publications for the industry that you are working in.

 

[sshep]

Thanks. This is an ethylene oxide plant. The operation will be continuous in non-TAR years, so on-stream time is expected to be more than 8000hrs per year (we hope).

I am really wondering if there is any reference for the standard of 8000 hrs/yr. It seems a relatively common number in process industries so there must be a reason, even if it is a convenience expression. Back when I did more economics we often used 360 day/yr for our ethylene plants, ect- i.e. 5 days/yr unscheduled production losses (98.5% availability).

 

[bimr]

8,000 hours of operation per year is widely taken as a benchmark of reliability and efficiency. Companies find it progressively difficult to squeeze out greater improvement when efficiency climbs above the 90% mark.

The number of 8,000 annual operating hours is based on the assumption of a continuous plant operation on seven days per week and 24 hours per day (three shift operation), corresponding to a plant availability of 91.3%

 

[EmmanuelTop]

There is no "standard" or "worldwide accepted" number of operating hours per year - this figure is completely driven by the required availability (or better say production efficiency = total volume of feed or product processed in a year divided by the required production in a year) of the new facilities, at one end, and project cost, at the other end. Increasing availability beyond a certain point will cause spending so much more money in order to achieve it, that the additional investment will never be recovered. So unless there are other factors in the play (i.e. criticality of production, security, or major safety issues), one normally stops increasing availability at a point where project NPV reaches the maximum.

Designing and building a plant for 8,500 operating hours per year in the case when the same plant is needed to operate only 7,000 hours per yer, is obviously a very bad practice. So everything needs to be seen from a "big picture" view.

Calculation of availability needs to take into account configuration and requirements of the upstream and downstream facilities. For example: if you have a unit that can operate at 105% or 110% capacity, and it has sufficient available storage volume for product(s), then you can easily calculate how many days per year the unit needs to operate at 110% capacity in order to accommodate for 30 days of downtime per year (and assuming that the facility downstream operates at 100% capacity all the time). This is only one example, and it may or may not be applicable to your plant. But there are many ways to work around the subject.



Dejan IVANOVIC
Process Engineer, MSChE

 

[EmmanuelTop]

See attached the glossary of reliability terms. I have extracted the pages from one of RAM reports.


Dejan IVANOVIC
Process Engineer, MSChE

 

[bimr]

Dejan, you are missing the point. sshep is buying a plant, not determining operating scenarios.

The basis of design for the process equipment assumes a production uptime of 90%. This sets the benchmark on the quality of the design and equipment. If you were buying a plant and requested a 50% uptime quality plant, you would be purchasing a plant with much lower quality and equipment.

Engineers typically assume 90% operating time for plants which work primarily with fluids, and 80% uptime for plants which primarily work with solids.

From Coulson & Richardson’s CHEMICAL ENGINEERING

"Time basis
No plant will operate continuously without shut-down. Planned shut-down periods will be necessary for maintenance, inspection, equipment cleaning, and the renewal of catalysts and column packing. The frequency of shut-downs, and the consequent loss of production time, will depend on the nature of the process. For most chemical and petrochemical processes the plant attainment will typically be between 90 to 95 per cent of the total hours in a year (8760). Unless the process is known to require longer shut-down periods, a value of 8000 hours per year can be used for flow-sheet preparation."

 

[EmmanuelTop]

I understood the OP is purchasing a licensed unit for which the licensor guarantees 8,000 hours uptime per year. If the required uptime hours were not contractually defined prior to purchase, there is not much he can do at this point. On the other hand, he says that this amount of uptime hours is exactly what is needed to achieve the required annual capacity. So maybe that is the only catch - the uptime hours were calculated backwards based on the required annual capacity, and then there are no problems at all.

Bottom line is that no licensor will provide guarantees for any extra capacity or extra operating hours, apart from what has been defined in the contract.

Maybe sshep can clarify his question, but it does not make any difference if you are buying a plant or designing one - as far as availability calculation is concerned. If there is a target availability that needs to be met, certain upgrades are required (adding units in parallel, change in configuration - if applicable, etc.). The only difference is who does the engineering for upgrades. For licensed units it is likely to have it done by the licensor. At an extra cost. The client pays for the extra equipment in both cases.

Latexman is correct when talking about refinery units. Most of them are designed to operate continuously for 3 or more years. That is quite away from any textbook values. The same applies for many other facilities. So I stick to my previous statement that there is no worldwide accepted number of operating hours which applies to all plants and all industries. It is a misconception.

Dejan IVANOVIC
Process Engineer, MSChE

 

[bimr]

"but it does not make any difference if you are buying a plant or designing one - as far as availability calculation is concerned:"

On the contrary, if you purchase a process plant and specified 50% availability, you would be purchasing a plant without redundancies such as spare pumps. The pumps would be on a shelf instead of as installed spares. There would be much more cost cutting in equipment as well. No standby power, etc.

If you purchased a plant and specified 8,000 hours of operation per year, you are purchasing a plant that is capable of operating 24/7.

A plant with 50% production uptime will have less cost than a plant with 90% production uptime.

Items like 8,000 hours are basic goals for a project when the project is sponsored and envisioned. The project manager is responsible for making sure that the project meets the goals.

Increasing operating efficiency results in less CAPEX and is more profitable.

 

[EmmanuelTop]

Items like 8,000 hours are NOT basic goals for all plants. This simply cannot be generalized. One gets 8,000 hours either if he specified this figure based on project requirement, or if he didn't specify anything and waited for the licensor to give him a figure they are comfortable with, while achieving daily/monthly/annual throughput.

Dejan IVANOVIC
Process Engineer, MSChE

 

[bimr]

In the English language "like" means "having the same characteristics or qualities as"

There is a difference between "Items like 8,000 hours are basic goals"

and "8,000 hours is a basic goal for all plants".

The OP's original question is where the 8,000 hour figure comes from and it was answered above.

 

[sshep]

Thank you my friends!

8000 h/yr comes from the licensor project design basis. The passage from Coulson and Richardson is a useful citation and evidence of reasonability. Much of the design is probably standardized by the licensor for plants all over the world, so choosing 8000 h/yr isn't so such a mystery. I am sure we will do better, but at least I can answer the questions the basis has generated. Thanks again.

best wishes always,
SShep

 

[mbt22]

Also look at whether the rate you want is annual or daily - an 8000 hr/yr 1,000 kte/ye plant will be about 10% larger than a 8760 hr/yr 1,000 kte/ye plant.

It's just a number for setting the hourly throughput, I suspect, not a guarantee or even indication of actual reliability. Trying to claim a specific reliability before detailed engineering is complete (and probably even then!) is a fools errand.

Matt

 

[zelgar]

Considering the maximum number of hours you could operate in a year is 8,760 hours, 8000 hours isn't too far off. Note, if you need permits for the operation, they may base estimates on operating 24 hours/day for 365 days or 8,760 hours per year.

 

[enderlord]

sshep,

If you are saying that the total annual capacity is the same as the one specified by you (as you say... "needed") than we could assume (as we do not know exactly) some of the reasons:
1. fit for purpose solution (lean and clean design)
2. not using redundancy if possible, thus eliminating extra equipment which would of course increase the availability of the plant, however with unnecessary costs (both CAPEX and OPEX)
3. again, less redundancy (sparing) also means that during the maintenance period, most of the equipment must me subject to maintenance since it can't be performed while spare equipment is online

As Dejan pointed out very well, there is no standard. Whatever you need to achieve as "availability of the plant" or "production availability" must be specified within the BoD or the "contract" itself, and that shall be the basis of the design, and you might want to subject the design to a RAM Study. Of course, the real test shall be done once the plant is operational, and having a couple of months to one or two years in operations.

I hope i have added some new information and not repeating the ideas which were already posted.

cheers,
vlad