I try to get a realistic number on ‘selection cost’ for people and travel involved in a DCS-selection process for an end user.
• ARC stated that an average DCS system cost around 1,5 million USD.
• Company’s use for a single between 1-3 people in a core team to do the selection. Corporation wide project can take up to 12 people full-time for 3 years.
• Yokogawa stated that the selection process will take between 9-24 month.

My first estimate (conservative) is that it will use at least the workings hours of 1 employee for a year and additional travel for visits.

To translate this to Euro /Dollars, it will be at least 100.000 euro or 150.000 USD.

Do you think this cost numbers are realistic or should it be higher or lower?

What are the selection cost for a DCS-system

Willem,

Without knowing the inhouse capabilities I would suggest that you break it out in "boxes" or like a flow chart. The flow chart could look like: field instrumentation, servers or APC's, IT (LAN, WAN), applications and integration. Once identified the needs you can then determine the people needed and its associated cost. It may be that some of the parts or part can be given to a trusted outside party and therfore reduce some inhouse cost and time. Another approach is to have a vendor selection process and evaluate their systems and technologies. And go from there. I do agree with Yokogawa's timetable, it depends on the control process and their compliance issues.

Regards,

Alfredo

Few Basic things such as How many I/O's the system has to Handle & the System Architecture you are going to implement to meet customers requirement.Things such as How many I/O modules,Communication modules such Profibus DP Module,No.of Client Stations,Remote I/O Panels,OPC Licenses do contribute lot in selection as well as costing.Time table (8-10 Months).People required.Ladder/FBD program-1 Expert,FrontEnd/Graphics-1 Person,Electrical-2-3 person's,Drives-1 Person,Process Consultant-1.

Willem,
In my experience, the most important step is to define correctly the enduser's needs and wants. This might take up to 6 months. Writing the bid spec and selecting the vendors for the bid is another step. The last step is choosing the offers evaluation criteria and rating the offers. I think 6 to 9 months should suffice for a 3 people team and the cost you mentioned is reasonable.

Regards,

Radu

Looks like all the companies with a strong meaningful standards based preferred supplier are saving themselves alot of money as well as shortening the leadtime significantly.

This does not always mean single source because some clients like the idea of rotating suppliers to keep the commercial deals nice and sharp and I guess share some risk just in case 'your pick' starts to go 'off-course' in the longer term. In reality if there is any window shopping it tends also to be done outside the selection zone.

My opinion is that the numbers quoted above are rather inflated for most situations. Many times there are 'unofficial' preferred vendors which shorten the process, the core team members are usually part-time and the initial specifcation can be rolled into the main contract in various ways and even performed by prospective keen vendors with good industry knowledge or direct experience of the end customer.

I think the cost you estimated is approximately correct for the 1st stage of DCS selection. For example, generating a short list of acceptable DCS vendors. That was my experience from leading selection teams corporate-wide for five years.
If you have to select one vendor, a more elaborate FSD needs to be generated and that eats a lot of time in a company especially if they don't have central engineering with full authority for this function.
The 3rd stage would be business in nature (a bid evaluation), that should be dependent on how many vendors have responded.
To give a short answer, I would double your figure (200-300 kUSD) to select a DCS vendor for a company / project / site.
I am not sure how much information I can share, but let me know if any…
Br/ wadei

Yeah I would agree with Wadei in that you would need a FSD or BOD. If you lack any of these, then you may want to look a your own spec generation with your core team. Your FSD/BOD covers all of the above. Also, there are engineering firms or companies that can help in generating this document.

The FSD costs are clearly separate in this discussion as these can vary widely depending on scale/complexity (upwards) or if the process is simple/small already defined with a copy-effect bonus.

I reckon there will be different rules of thumb for DCS selection costing depending on the industry verticals ie pharm/oil/gas/chemical/energy etc.

We have a lot of good comments here. An additional way to look at this is to consider the scope. If we're talking about a project involving DCS selection, programming, engineering (including instrument, electrical, and civil/structural), and construction, then I would expect this phase to be about 1% of the estimated total installed cost. For a small plant with a few thousand points (and depending on scope, of course) , you could be at around $25 million for DCS, programming, engineering, and construction, so DCS selection plus some front-end engineering (scoping and cost estimation) could run about $250,000 or so.