7 replies to this topic

[Eprocess]

A few number of chemical plants are to be constructed in a complex, and they are going to get utilities from a master offsite facility. I was wondering what is the right approach to determine the offsite facility`s capacity for several services (such as steam, DM water, Instrument air, and...).

Every plant it the complex has its own consumption pattern, peak flow and daily consumption for each service. What is the reasonable practice for determining the optimum capacity offsite facility`s ? What kind of data should be in hand except what I mentioned?

Should we be conservative and design the system based on peak flow of every service in each plant?
or
We should take daily usages to the account, and consider a coefficient of usage?


Thank you.

[ankur2061]

Eprocess,

I would consider the following approach:

1. For each plant / chemical complex, get the utility consumption figures (steam, cooling water, chilled water, service water, raw water, nitrogen, demineralized water, fuel gas etc.) as provided by the technology licensor of the plant / chemical complex.

2. For open technologies utility consumption per unit production of the chemical are available freely from various resources. Based on the unit production rate, calculate the total utility consumption (each type) for the nameplate capacity of the plant.

3. Total up the various utilities as mentioned in 1 & 2.

4. Based on the figure arrived from 3 provide a design margin of say 30% on the calculated figures for each utility except water. Refer 6 & 7 for calculating water requirements.

5. Use the figure as calculated in 4 to fix the design capacity for each utility generation unit.

6. For raw water intake / pumping, you may avoid adding design margins at each water utility. Estimate the various water requirements (cooling water make-up, chilled water make-up, demineralized water consumption, service water etc.) for each of the units. Total up under the individual water categories for all plants/ complexes without any margins. Then total all the water categories to get the total raw water requirement without adding any margin. Now put a design margin of say 30% on this total to design the raw water intake and pumping system.

7. Alternative approach for water utility would be to provide a design margin say 30% for each water utility at each plant / complex and add this up to arrive at a total figure for all water. This figure may then be directly used to design the raw water intake / pumping without adding any design margin.

Hope this gives some idea about how to proceed in designing a master off-site utility complex.

Regards,
Ankur.

[Eprocess]

Thank you Ankur,

The problem is for intermittent consumption, where for instance a considerable amount should be supplied in short cycles.
Every plant has buffer tanks for some services, should I count on them to mitigate the effect of intermittent peaks or they are just provided as a back-up in case of temporary shutdown in off-sites, and should not be depleted in normal operation?
For instance if the design flow rate of steam for a plant is 10 ton/hr and the daily usage is 120 ton, should we consider the daily usage in off-site design, and reduce the capacity or not.

I mean because several plants are being fed, so maybe we can reduce the system size by studying the consumption patterns, instead of just summing them up.

[ankur2061]

Eprocess,

Normally the design margins that are provided for the utility generation unit capacity should take care of the short-term or intermittent peak rates. The design margins serve another purpose also. During subsequent debottlenecking of any plant, these design margins on utilities come in handy and there are no concerns for designing a new utility generation unit as long as these margins can cater for the increased utility consumption subsequent to the debottlenecking.

However, in certain circumstances deciding the intermittent peak utility consumption may be purely judgemental and based on past experience of the design engineer for a similar kind of plant / unit.

My personal experience when preparing plant utility summaries has been that the technology licensors generally provide a decent margin for each utility consumption in order to protect their interest of demonstrating guarantee figures for utilities during the plant commissioning and guarantee run.

Hope this takes care of your concerns about intermittent peaks for utilities consumption during the operation of a plant.

Regards,
Ankur.

[tarafdar]

The following points may be considered also :

1.During designing utility facilities plant operation flexibility and reliability are to be considered.
Say if steam demand is 100mt/h,we may go for a 120mt/h boiler.But it will be better to choose two boilers having capacity of 60mt/h .It will increase plant reliability with increased capital investment & running cost.With one boiler your whole process may have to shut-down in case of boiler failure.But with two boilers ,you may be able to keep some sections running during trippage of one.

2.For DM water ,DM storage capacity to be considered .

3.For instruement air ,instrument air reservoir at high pressure may considered.

4.You have to thoroughly study different plants start-up & shut-down procedures for proper utility design.

Thanks

Edited by tarafdar, 19 October 2011 - 09:20 AM.

[fallah]

6. For raw water intake / pumping, you may avoid adding design margins at each water utility. Estimate the various water requirements (cooling water make-up, chilled water make-up, demineralized water consumption, service water etc.) for each of the units. Total up under the individual water categories for all plants/ complexes without any margins. Then total all the water categories to get the total raw water requirement without adding any margin. Now put a design margin of say 30% on this total to design the raw water intake and pumping system.

7. Alternative approach for water utility would be to provide a design margin say 30% for each water utility at each plant / complex and add this up to arrive at a total figure for all water. This figure may then be directly used to design the raw water intake / pumping without adding any design margin.

Ankur,

I think there is no difference between 6 and 7 methods as above, and both methods would arrive the same final value for design the raw water intake. Actually, in 6 you apply design margin at first and in 7 it would be applied at the end. Am i wrong?

Fallah

[ankur2061]

Fallah,

Yes, you are right. There would be no difference between the figures from 6 or 7. Thanks for pointing this out. I did not do my math prior to writing this.

Regards,
Ankur

[kkala]

The original question is whether peak consumptions for each separate plant should be summed as they are. Or each peak consumption should be multiplied by a coefficient of use and then summed. There may be other ways too. Utility capacities have to be early specified in project life to place the orders in time.
I would say that generally maximum consumptions at normal operation should be summed, plus some extra margin (Spem); besides storage capacity should be taken into account. This is understood to comply with suggestions in the forum, let us investigate each utility, offsite, or similar consumption (*) individually. Following thoughts are influenced by understood local refinery practice, including a recent medium size upgrade (new Vacuum distillation+Hydrocracking+Hydrogen units+utilities+offsites).
1. Demineralized water: Spem. However part of it is substituted by treated condensate, whose available supply can have variations. Demi water (DMW) production plant will have to stop for maintenance / cleaning too. Thus DMW storage capacity should cover this, and DMW plant should cover any reduction in condensate supply.
2. Nitrogen: Spem, concerning gaseous N2 production. Liquid N2 to storage (to pass from evaporators) should ensure coverage of peak local demands of the units. During commissioning N2 supply "stations" fed by trucks may be also needed.
3. Raw water: Spem, storage covers the case of shortage.
4. Plant air: Spem (there is an opinion that it is never enough, used in tools, cleaning, etc).
5. Instrument air: Spem, 10 min buffer storage to each unit for case of supply failure.
6. Fire water: Spem not applicable. Capacity should cover the single fire area of maximum fire water demand ( factory is divided into numerous fire areas, per NFPA).
7. Gases to flare: Flare load is defined through a special study to estimate how many of potential releases can occur simultaneously. However generous margins on this result would allow for future expansions, which has been very fruitful here.
8. Steam: Max consumptions of all consumers is summed for several realizable scenarios, and maximum of the sums (plus margin) gives the generation capacity for that level of steam. In mentioned upgrade two boilers can cover the needs, three are operating at reduced capacity.
9.1 Fuel oil to boilers: Intermittent transfer to daily local tanks, filled in one shift. Thus transfer rate ~3 x fuel consumption corresponding to maximum total capacity of 3 boilers.
9.2 Fuel gas to boilers, alternative fuel: Corresponding to maximum total capacity of 3 boilers.
10. Electricity: Having heard of coefficients of use in the past, now a procedure like that for steam seems to be followed.
Note: During factory start up, units are successively put on stream. I assume utility consumptions are satisfied in this way. In case of an excessive utility demand for a specific reason, this has to be covered capacity wise (not heard of such a case, except N2 for commissioning).
Above examples merely indicate a sort of approach that can be applied during basic engineering, trying to avoid confusion later. Revision would be made in detail engineering (probably not, if needs are covered), hopefully to a limited extent. It is not an easy task and Engineering Contractor has implemented it here, concerning the mentioned upgrade.
(*) consumptions not considered as utilities were also reported.

Edited by kkala, 20 October 2011 - 10:42 AM.