Free estimating resource

Risk Analysis in Estimating: do-it-yourself (DIY) Monte-Carlo Simulation

Vikas KhaitanLeave a comment
5–7 minutes

No matter what class of estimate it is, every estimator makes his or her slightly different assumptions based on their own experience. An estimate is finally an opinion of a particular estimator. Some estimators tend to be pessimistic in every assumption, factor, historic data they use, ending up with a very conservative (i.e. high) overall estimate. For the same technical inputs, which might themselves not be perfect, another estimator might make optimistic assumptions for individual elements of the project, which might make the overall estimate very low. In both cases, the estimator is likely to add a contingency to their estimate.

Most often, the assumptions made by individual estimators are not transparent as they are generally not properly recorded or documented. Besides this, the decision-maker is usually given only one estimate from an individual estimator to look at, which does not give them the option to question or critique the possible range of the overall estimate.

If somehow the same estimate could be done by multiple estimators, and all those estimates were tabulated together, we would get a range of estimates, with a maximum and a minimum value, which would allow everyone to understand the possible range of cost outcomes of a future project. But it would be very expensive or nearly impossible to get the project estimated by several estimators at the same time.

But there is a workaround to this! We could instead use MS Excel to do multiple estimates. And not only that, but we could also get Excel to run 5,000 or even 10,000 estimates almost instantaneously (for free!!).

  • This is like predicting 10,000 possible scenarios out of the infinite possibilities (simulating reality by trying to analyse many possibilities).
  • If we could analyse and simply present these 10,000 results, we could in a way say what the likelihood (probability) of success for a particular point estimate would be.
  • This would be better (and more transparent) than trying to present a single number.

Nobody knows the future, so whatever we predict is likely to be wrong; the point is to reduce the amount of error in our predictions. So, instead of trying to produce a correct single number (Point Estimate), I suggest that we generate a reasonable range instead (Range Estimate).

This is called Monte Carlo simulation, and it is a mathematical model of probability analysis. There are various companies selling risk analysis (Monte Carlo simulation) software in the market, but the cost for those tools (toys!) are sometimes not justifiable even within large organisations. But a range estimate would be really useful by allowing decision-makers to understand what contingency would be suitable for a particular estimate and at what risk level are they pitching the overall number.

I have developed something easy and simple to help bring the team together and run these simulations to see what the possible range of the overall estimate could be:

  • Let us say that you come up with a point estimate for the project – say the total is 100
  • And the estimate is built up by adding, say, 20 different items (or more)
  • But each of those different items could have been assigned a different number (within a particular range)
  • The team can come together to understand the assumptions made by the estimators for those individual items, and can together decide an appropriate range for each item

Once those ranges are defined, we could ask Excel to randomly pick any number between those ranges for each of those items (by using the Excel formula “Randbetween”) and add them up to come up with a new estimate, as shown below:

We could do this 10,000 times to come up with 10,000 new estimates. This is how a simple DIY Monte-Carlo simulation model could be made.

Typical results will look like the following:

Interpretation:

  • The first graph plots all the 10,000 estimates (simulations) generated by this method [using a simple scatter graph]
  • Based on these 10,000 simulations, the second graph shows what are the chances of completing a project within the “Point Estimate” value. [this simply plots the Percentiles of all the 10,000 values using a scatter graph]
  • Do note that in a real life situation there is no chance of all possible worst cases happening simultaneously, nor all the best cases can happen for a project. Therefor the total of all the maximums and all the minimums does not make any sense, and the Monte-Carlo results also show that they never occurred within the 10,000 simulations.

The file attached below is the model I have developed as explained here and is a free estimating resource for anybody who wishes to download and use.

20250316_DIY_Monte_Carlo_forEstimating_Rev0 (free resource)Download

This example template file is designed to take up to 50 items in the inputs tab (this is the breakup of the point estimate), with their estimated values and the max-min ranges for each value. The inputs tab is then linked to the Monte-Carlo Results tab where the 10,000 simulations are run by simply pressing “Fn&F9” together on the keyboard. This tab also presents the results in the graphical form presented earlier. If needed, this file can be modified easily to increase the number of items adding up to a total point estimate.

Notes:

  • I have assumed that all the elements are mutually exclusive cost items and there is no interrelation between them.
  • I have used uniform distribution to keep the model simple. This means, once the ranges are defined, all the numbers within that range will have equal probability of occurring. Of course, this is never the case in real life, but it works as a simple model.
  • Here is where discussion and decision-making come into play – the team can sit together and understand the assumptions behind the various elements of the project and see what the likelihood of the overall number is if a different number is chosen within that range for that item.
  • This model will then help the team to understand where the overall point estimate is pitched at, and backup the contingency they want to add to the overall number for any class of estimate.
  • Other tools in the market are more sophisticated with triangular or other probability distributions, interrelated elements etc. which do make them more complicated and difficult to explain, and in my opinion, they may not add that much more value to decision making [as they could be perceived as black-box by decision-makers].
  • This simple DIY tool can be a useful tool for initial estimate reviews and discussions.

ENJOY!

Biomass Power Plant Benchmarking (Open-Source Example)

Vikas KhaitanLeave a comment

In one of my past blogs, I developed and shared an open-source benchmarking example for gas and coal fired power plants. I have now developed another open-source benchmarking graph, using internet-based data, for a typical biomass power plant project. I have mostly used news articles giving the total value of the contract and capacity in Megawatt (MW) of the power plant.

20240622 Biomass Power Plant Benchmarking (Open Source)-Rev0 (free Resource)Download

The data is tabulated (in the attached Excel file) with links to the various websites for cross-checking purposes. Within the file I have also escalated the contract values to the current year using a nominal escalation factor (which will vary depending on the market conditions). This can be modified to any future year when needed. This gives a feel of what the market has paid for a typical biomass power plant. These are not out-turn costs but initial awarded values and thus should be treated as such. This is not exact but gives an idea.

No two projects are similar in scope, but what this does is to give an opportunity to the reviewers to understand any special / specific requirements of a project which might make the current project estimate different (like additional fuel conditioning, remote location, additional transmission lines, piling requirement, etc.). The overall plant cost will also vary based on the technology used, quality of fuel used, efficiency of the plant etc.

Also to be noted is that this is only an EPC cost and any Pre-FEED, FEED, other Owner’s costs, service agreements, operation & maintenance costs are to be added separately as needed.

This can be used in addition to or in the absence of any in-house or any third-party benchmarking data. This can be easily shared with anybody and can be used as a cross-check of the detailed biomass power plant estimate at a $/MW level.

The graph is also suitable for coming up with quick order of magnitude estimates for bid / no-bid analysis, initial project sanctions, bid evaluations, etc.

The graph shows the typical trend of reducing unit cost for larger capacity plants.

I have just done some research to show what can be possible with freely available data and so the graph is not considered comprehensive and can be enhanced with more data points if further research is undertaken. Location / area specific graphs can also be generated if enough data points could be gathered.

Feel free to use this as you see fit with or without modifications.

Special Notes:

Note1: Various studies show that depending on how the estimate was performed, the actual out-turn cost could potentially be much higher. For this reason, note that in this graph, I have also indicated a 75% cost overrun (as an example) to show how the actual out-turn cost might look like. This percentage can be varied within the Excel file to see various sensitivities.

Note2: I have also plotted some additional data points on the graph using cost data published by IRENA (Renewable Power Generation Costs in 2022), suitably escalated to the current year. I have used the high data for North America (see page 181 of the IRENA report) as an example for comparison with my internet based open-source data. I am suggesting that this can be used as another cross-check.

Note3: For some projects, O&M for number of years was included within the award value, which I have subtracted (using a nominal O&M cost per year). This can also be modified as needed within the Excel file.

Tank Farm Benchmarking (Open Source Example)

Vikas KhaitanLeave a comment

I have developed an open source benchmarking graph, using internet-based data, for oil storage tank farm projects. I have mostly used news articles giving the total value of the contract and total storage capacity in cubic meters (cu m) of the tank farm.

20210818 Tank Farm Benchmarking (Open Source)-Rev2Download

The data is tabulated (in the attached Excel file) with links to the various websites for cross-checking purposes. Within the file I have also escalated the contract values to the current year using a nominal escalation factor (which will vary depending on the market conditions). This can be modified to any future year when needed. This gives a feel of what the market has paid for a typical tank farm of a particular total capacity. These are not out-turn costs but initial awarded values and thus should be treated as such. This is not exact but gives an idea.

No two projects are similar in scope, but what this does is to give an opportunity to the reviewers to understand any special / specific requirements of a project which might make the current project estimate different (like remote location, additional piping, additional pumping, additional pipelines, piling requirement, etc.).

Also, to be noted is that this is only an EPC cost and any Pre-FEED, FEED, other Owner’s costs, service agreements, operation & maintenance costs are to be added separately as needed.

This can be used in addition to or in the absence of any in-house or any third-party benchmarking data. This can be easily shared with anybody and can be used as a cross-check of the detailed tank farm estimates at a $/cu m level.

The graph is also suitable for coming up with quick order of magnitude estimates for bid / no-bid analysis, initial project sanctions, bid evaluations, etc.

The graph shows the typical trend of reducing unit cost for larger capacity tank farms.

20210818 Tank Farm Benchmarking (Open Source)-Rev2

I have just done some research to show what can be possible with freely available data and so the graph is not considered comprehensive and can be enhanced with more data points if further research is undertaken. Location / area specific graphs can also be generated if enough data points could be gathered.

Feel free to use this as you see fit with or without modifications.

Special Note:

This analysis could be used to benchmark the complete project cost of any site-built tank.

Sometimes, site-built tanks are included as part of larger projects. In those estimates, the site-built tanks are considered as part of the process equipment items and are typically costed as part of the equipment list. In those situations, it would be a mistake to include the total site-built tank project cost (as per this benchmarking analysis) within the equipment cost account. I would suggest including only the steel supply & prefabrication cost for the tank as part of the equipment cost, which would typically be around 25-30% of the overall tank farm cost (as benchmarked here).

Apart from the steel supply & prefabrication, the other costs which make up a complete tank farm project include all civil works including bunds (if required), installation costs, and the costs of supply, fabrication and installation of all the associated piping, electricals, instrumentation and control items. Finally, costs such as shipping, spares, engineering, project management, site mobilisation, construction management, commissioning and any risks and mark-ups are also part of the overall tank farm project cost.

These additional costs are generally added to the estimate of the whole plant and is not separated out for the site-built tanks within the plant.

Weight Tables for Quick Piping Material Estimate

Vikas Khaitan1 Comment

In one of my previous blogs, Quick Piping Material Estimate, I provided indicative benchmark $/kg rates for various carbon steel piping items, and discussed how these rates could be used to quickly estimate piping material costs.

To be able to do that, the estimator would need access to the weight tabulations for various piping items, which I mentioned can be easily obtained from manufacturers of pipes, fittings, flanges and valves.

The file attached below has tabulations of different piping items which I have collected from various manufacturer catalogues. This is a free estimating resource for anybody who wishes to download and use.

Piping Weight Tables (Excel file) – Free Estimating Resource

The weight tables available are for below Carbon Steel piping items:

  • Pipes
  • Fittings
    • 45° Long Radius Elbows
    • 90° Long Radius Elbows
    • Caps
    • Tees
    • Reducers
  • Flanges
    • WN Flanges
    • Blind Flanges
    • Spectacle Blind Flanges
  • Valves
    • Globe Valves
    • Check valves
    • Gate Valves
    • Full Bore Ball Valves
    • Reduced Bore Ball Valves

Weight tabulation of additional piping items can be added to this file if required. Any missing weights for particular sizes can be prorated, from nearby weights, for estimating purposes.

The file also shows how these weight tabulations can be used in conjunction with Excel functions like HLOOKUP and VLOOKUP in order to quickly list out the weights of the various items. This is demonstrated in the “Piping Material Estimate Template” tab within the file.

The following inputs are required to auto-read the weights from the various tables:

  • Item description (pre-defined)
  • Sizes (in inches)
  • Schedule for pipes and fittings (e.g. 40, 80 etc.)
  • Pressure rating or piping class for flanges and valves (e.g. 150#, 300#, 600# etc.)

The file shows how the inputs should be tabulated and listed to make this work. The formulae can be repeated for thousands of rows of MTOs helping to get the weights within minutes. And using the $/kg rates, the whole piping estimate can be done within a few hours (or even minutes), rather than weeks.

In case of resource shortage, this tool can be used as an effective estimating method; it can also be used by the reviewers to check if the piping estimate done by other more time-consuming methods are in the right ball park. Moreover, a blank inputs table can be shared with the piping engineers to make sure that the inputs come in this prescribed format, thus reducing the time required for any re-formatting.

I have been successfully using this method for many years and would suggest that this is one of the ways the estimating community can increase the team’s productivity and provide better “value for money” by substantially reducing the time and cost for estimating.

Note: These are standard product weights from some manufacturers. Products from specific manufacturers could vary slightly in weight. If the same weight tables are used to calculate the $/kg rate than these slight variations will not make any difference.

Also note that if the design requires non-standard sizes, then the weight of those piping items would need to be calculated from first principles.

Quick Piping Material Estimate

Vikas Khaitan6 Comments

When estimating any oil & gas project, detailed piping estimate is one of the most time-consuming elements. Piping material take-offs can sometimes be in the form of several hundred or even thousands of line items, which an estimator takes weeks to estimate. Some companies have specific piping estimators for the job.

In all oil & gas projects, there are several pipes going in and out of the various equipment items helping the process and utility fluids to move around in the required direction and destination. This piping forms a major portion of the final plant and the cost estimate. There are several piping items such as: pipes, fittings of various kinds, flanges of various kinds, valves of various kinds, and the corresponding gaskets and fasteners. There could be several sizes and pressure ratings of the various piping items. And then there may also be various metallurgies involved. The material take-offs are generally done separately for various lines, which means, similar items could be repeated several times. The whole list is generally very long by the time it comes for estimating.

The piping estimator then tries to find the cost of individual line items from the in-house database and invariably takes a long time to find an exact match for the material, size and schedule of a specific piping item. If the company has previously used similar items in another recent project, then there might be a possibility of getting some in-house data, but even then, it might not be available for all items. The team might end up factoring those items from another nearby size and/or schedule. This whole process is very laborious. Even with databases and software, no company could have an exhaustive databank of prices of all possible piping items.

Depending on the budget and resources available, the team might decide to go out to the market to get recent prices for all the piping items, but even that takes a considerable amount of time.

Once this laborious process is completed, it becomes very difficult to review the overall estimated piping supply cost, as it is nearly impossible to go through each line item.

I am going to suggest a surprisingly simple method to help expedite the piping estimate and review process. For this, all piping items can be converted into weights and then multiplied with an average $/kg rate for the various types of piping items to arrive at a rough piping material supply estimate for comparison.

The table below offers indicative benchmark average $/kg rates for various carbon steel piping items (free estimating resource).

You can choose to use company specific rough $/kg rates for various piping items (instead of the above rates), which can be derived by using database information from past projects or previously obtained budget quotes. This is preparatory work, which can be done when there is less load on the estimating team. Similar rate tables can also be developed for other material such as stainless steel or duplex steel, or factors can be developed to modify the $/kg rates from carbon steel to the higher grades of material. (I will try to post indicative factors as a future blog.)

The weight tabulations for various piping items can be easily obtained from manufacturers of pipes, fittings, flanges and valves (I will share the piping weight tabulation in another post).

This method helps in quickly estimating a rough number for the material cost of all piping items. This can either be used when there is insufficient time or estimating resources to complete a more detailed estimate, or used by the reviewer of the estimate to cross-check the piping estimate done by others. If the weight tables are properly set up then this method can help arrive at a rough piping estimate within hours (or even minutes if the inputs are provided in specified formats), against weeks of work that is generally undertaken.

Note: The benchmark rates above are indicative only and intended to demonstrate how this method would work; they can help arrive at a reliable overall piping material estimate when time is short. 

Piling Costs: Rough Estimate Model for Benchmarking

Vikas Khaitan1 Comment

Once when trying to benchmark piling costs with available information from various in-house projects, the data showed an apparently inexplicable variation in the $/m rate for the piling works. In some cases, it showed 2-3 times the $/m rate for the same diameter, and in other cases, the data showed similar $/m rate even though the diameters varied between projects.

After the initial failed benchmarking exercise, I started considering the technicalities of piling and understood that there are different kinds of piling specified for various projects depending on different soil and project conditions. I realised that I was making the mistake of comparing the $/m rate of different piling types used in different projects. This wrong comparison had created the above-mentioned variations in the overall rate.  At this point, I separated the available data into the various types of pilings. That was a good start, but then I did not have enough data points for the individual piling types to be able to generate any sensible benchmarking curves.

But benchmarking was needed for the estimate approval process.

To meet this need, I developed a small estimate model to calculate the rough $/m cost for various types of piling and for various diameters. I assumed an indicative length of pile to generate the graph. I used indicative material and labour rates, norms and productivity factors to calculate the cost of the various types of piling. I included an indicative hiring cost of piling hammer or rig as required for a specific type of piling work. All these inputs and assumptions could be modified for specific project and location to help generate a customised benchmarking graph. I have also assumed a total number of pile for this specific example and would caution that the per meter rate could substantially vary if the total quantity is significantly different. But this still gives an idea and relative cost differences

The attached file shows the proposed working.

Piling Rough Estimate Model for Benchmarking-Rev0 (free resource)

This shows how the costs could possibly vary with diameter and piling type and can be used as an in-house estimated benchmarking graph. Any past project data (if available) could be superimposed on this graph for easy comparison. Any contractor bids can then be compared against this graph to demonstrate the reasonableness of market pricing. This can also help the engineers choose and compare more than one piling type, if technically suitable, to use in a project.

Design Option Selection Estimate Layout

Vikas KhaitanLeave a comment

During the initial phases of any project, there are multiple technical solutions that the team could choose from for the overall design or for individual technical details. These various options need cost comparisons for effective selection and decision making. It sometimes is a challenge for the team, to get everybody involved to agree on a single solution. But a proper layout and presentation of these estimates could help to ease the understanding and in the decision making.

When presenting an option selection cost estimate, it is very important to tabulate the varying scopes (at a high level), alongside the costs to transparently present the differences in scopes between the various options. I would also recommend presenting all the options in a single tabulation if possible. This helps in easy comparison of the scope differences and then to see how the overall cost varies.  A single tabulation would help during any discussions with the technical team and might highlight any items that could have been missed or not defined correctly when compared to each other.

Also, the unit cost basis for any individual item should not vary between the options. Any variation in unit cost should imply a new item description. The only difference between the options should be the quantity of individual items. Keeping the same unit cost basis gives the confidence to the decision makers that they really are choosing between different technical options. Do note that there could be specific items in an option that may not be required for other options.

Attached is a dummy template that could be used for any option selection estimate. I have personally used similar tabulation for various study estimates and successfully helped in the decision-making process.

Design Option Selection Estimate template (free estimating resource)

Case in point:

In a particular project, involving a wide range of technical solutions, various options were estimated and the cost summaries of the individual options presented separately. This did not easily allow the decision makers to understand the technical differences between the options. When reviewing the cost estimate, most of the time was spent in understanding the technical differences and the reasoning behind them, as it was not easily obvious why the costs of the various options were so different, and what benefit did the project get by investing more money. The technical scope and costs being separated, did not lend itself to easy decision making and no decision was taken. It took the team months to select an option.

I have been practicing presenting design option estimates in a single page layout and have most of the times successfully convinced the team to easily understand the various options technically and commercially. This has always been helpful in decision making and sometimes only within a single meeting.

Note:

Not all items that are included in an option selection estimates can necessarily become part of a single project. For instance, when deciding between two plot spaces for a process plant, the road that needs to be built could be of different length in each option and might help decide which is a more favourable overall cost project, but the actual road might not be included in the project scope at all.

This example shows that all scope differences, even beyond the current project boundaries, should be considered when trying to select between various competing designs and projects.

Thumb Rules for Engineering Costs

Vikas Khaitan5 Comments
2–3 minutes

Sometimes a very rough high level engineering cost for a future project might be needed for budget approvals, initial planning, contracting strategy discussions, benchmarking of bid prices or negotiating with the engineering contractors. The rough cost could also be helpful to plan the manning level in the design office.

As a thumb rule, for most Oil & Gas projects, typically, the engineering costs generally ranges between 5 and 15% of the overall CAPEX (Capital Expenditure), depending on the scope of engineering required, project size, complexity, brownfield / greenfield development etc.

The percentage will be at the lower end for a larger project (higher CAPEX) and will tend to be at the higher end for smaller CAPEX projects.

The various engineering scopes that could be included in any one or more engineering contractor’s scope of work could be as follows:

  • Concept work
  • Pre-FEED engineering
  • FEED engineering
  • Detailed engineering
  • Procurement services
  • Follow-on engineering
  • Site survey works

The below tabulation shows typical indicative engineering costs for various projects with varying CAPEX values

20170218-enginnering-cost-thumb-rules-rev0

Engineering Cost Thumb Rules (Excel Version) – free estimating resourceDownload

These rough percentages are very much high level indicative numbers, based on experience, to just help with a quick calculation of the total engineering costs. The table above can be utilised to calculate the indicative high level cost for the engineering effort required for a scope (for example, only FEED engineering, or only detailed engineering etc.).

The actual engineering cost for individual projects will depend on the project complexity, location of engineering, available skill level, previous experience with the technology and many other factors.

Ideally, of course, deliverables and activity based resource requirement would be established and costed by Individual estimators / project engineers to arrive at the overall engineering cost for the required scope. The high-level percentages suggested in this post can still be useful in such a case; they can be used as a rough guide to review, cross-check and critic the detailed calculation.

Note: The actual percentage for all the above scope of work could sometimes be much more than 15% of the overall CAPEX for small projects. The percentage could end up being less than 5% for very large CAPEX projects, specifically those which have several repeatable portions which make the total value of the project high, but would not require multiple engineering effort.

View Paid Services
New! Check out what is Flextimate™

Labour Productivity Factor Calculation Tool

Vikas Khaitan2 Comments

During the estimating process, direct labour hours are first calculated using estimated quantities from material take-offs and estimating labour norms like Gulf Coast, DACE or in-house. Calculated norm hours are for ideal working conditions not encountered in reality and only used as a starting point. The norm hours are then multiplied by a theoretical site specific Productivity Factor to arrive at the total on-site labour hour estimate. These site hours are then subsequently used to arrive at the total construction cost. The total site hours are also utilised for labour manpower planning and resource allocation.

The Productivity Factor is purely an estimating parameter and not something that is recorded during any project execution. It is a theoretical location / site specific factor used to convert the norm hours into estimated site hours. It generally depends on the site’s climatic conditions, soil conditions, water table level, permit control requirement etc.

As the ideal scenario, the estimating community needs location / site specific productivity factors for each standard estimating norm (Gulf Coast, DACE etc.). But productivity factors are not easy to arrive at as they could vary between contractors depending on issues like previous experience, engineering level and type, site survey quality etc.

It would be very convenient for every contractor to have their own historical databases of Productivity Factors for use in future estimates. To be able to correctly do this, contractors should have recorded the total direct labour hours actually spent on various projects along with a catalogue of the corresponding installed quantities for various equipment and bulk material. Only then a further exercise can be carried out to calculate the total norm hours based on the current estimating methodology. The actual total hours can be divided by the calculated total norm hours to arrive at the productivity factor for the historical projects to be used as a benchmark for future estimates.

However, most companies lack this kind of historical information gathering and analysis to support future estimates and even less for any specific country / region. Even if suitable historical data were available, such an exercise would involve a considerable amount of estimating effort, which may not be justifiable.

Most estimators finally end up making judgements to determine the best possible Productivity Factor to use for any particular site. As a result, when presenting to the client, any backup benchmark data is generally non-existent and both the contractor and the client teams try to convince themselves of the right factor based only on anecdotal experience.

To help create and substantiate the factor to be used in any estimate, I have created a Productivity Factor Calculation Tool in the attached file.

Productivity Factor Calculation Tool-Rev0 (free estimating resource)

This is based on factors, explanations and comments which I have collected over the years, and the product of innumerable discussions during estimate reviews with different stakeholders.

In the tool that I am offering, I have listed 27 different factors that may affect productivity. For any individual factor, the project specific effect could range between none, low, medium, high or very high. Choosing any of these will affect the proposed productivity factor by a certain value. For example, the distance of the construction site from the population centres will affect the transportation time of the local workforce thus affecting the productivity factor to be used. For longer distances the user of this tool can choose “high”, which would then assign a higher additional factor to be applied to the norm.

I have in effect chosen an arbitrary value for each of the ranges based on experience but they can be modified by individual estimator in discussion with site teams. Assigning a number, even if arbitrary, allows a systematic approach to be taken, which is a better substitute to inconclusive discussions. This approach will allow a more focussed discussion with proper reasoning allowing the team’s experience to be better used and channelized. Estimating finally requires a number to be produced, and through my years of experience, I have formulated this approach, which, although imperfect at the start, has the benefit of focussing discussion and convincing the broader team, as well as the client.

This tool could also be used to calculate different productivity factors for a greenfield or a brownfield project at the same location.

Note that this is just an example working for a typical greenfield onshore oil & gas project, and needs to be modified to suit the specific region / country / project site. This should only be used as a guide and is not considered perfect or exhaustive. Additional factors could be added if deemed suitable. The presentation and relative effects / additional factors are my own and do not have any industry standard / basis.

This working is suitable to present as a backup to any estimate and clearly tabulates all the elements included in the proposed Productivity Factor. This also helps the team to understand why the total direct hours in any particular project are higher than for example another simpler site.

This tabulation, along with any reasonable historical data that can be obtained, will help to make the estimates more transparent and generate confidence in the estimates produced.

This article was published as an opinion piece in May-2018, in the Project Control Professional which is the journal of The Association of Cost Engineers.

Power Plant Benchmarking (Open Source Example)

Vikas Khaitan2 Comments

I have developed an open source benchmarking graph, using internet based data, for a typical power plant project. I have mostly used news articles giving the total value of the contract and capacity in Megawatt (MW) of the power plant and the type of fuel used. I have done it for only gas and coal fired power plants. This can be extended to other types as well.

Power Plant Benchmarking (Open Source)-Rev0 (free resource)

 

The data is tabulated (in the attached Excel file) with links to the various websites for cross-checking purposes. Within the file I have also escalated the contract values to the current year using a nominal escalation factor (which will vary depending on the market conditions). This can be modified to any future year when needed. This gives a feel of what the market has paid for a typical power plant. These are not out-turn costs but initial awarded values and thus should be treated as such. This is not exact, but gives an idea.

No two projects are similar in scope, but what this does is to give an opportunity to the reviewers to understand any special / specific requirements of a project which might make the current project estimate different (like additional fuel conditioning, remote location, additional transmission lines, additional fuel pipeline, piling requirement, etc.). The overall plant cost will also vary based on the technology used, quality of fuel used, efficiency of the plant etc.

Also to be noted is that this is only an EPC cost and any Pre-FEED, FEED, other Owner’s costs, service agreements, operation & maintenance costs are to be added separately as needed.

This can be used in addition to or in the absence of any in-house or any third party benchmarking data. This can be easily shared with anybody and can be used as a cross-check of the detailed power plant estimate at a $/MW level.

The graph is also suitable for coming up with quick order of magnitude estimates for bid / no-bid analysis, initial project sanctions, bid evaluations, etc.

The graph shows the typical trend of reducing unit cost for larger capacity plants.

20160425 Power Plant Benchmarking (Open Source)-Rev0

I have just done some research to show what can be possible with freely available data and so the graph is not considered comprehensive and can be enhanced with more data points if further research is undertaken. Location / area specific graphs can also be generated if enough data points could be gathered.

Feel free to use this as you see fit with or without modifications.