Risk metalanguage: how to make a proper Risk Register

One of the key deliverables in project risk management is the risk register.

The idea is that after risk identification, qualitative and quantitative analysis, and planning of risk response the tender manger (or project manager) should end having a list with all the risk identified at that point in time.

I assume that the content and format of the list will vary depending on the type of project and organization. However I believe that it should include at least:

  • Risk description
  • Impact on objective
  • Risk response strategy
  • Risk owner

The Project Management Institute add to this list other concepts like category (the taxonomy of risks always looked somehow arbitrary to me), probability (even more difficult to define with accuracy in many situations) and cause.

If you are familiar with the topic you will remember that, if a risk materialize, it will be moved from the risk register to the issues (or problems) register.

And here the interesting parts – I have been reviewing some risk registers that I have done in the past noticing that the description of risk was mixing three different concepts – cause, risk and effect.

“Presence of strong Workers Unions in the area”, for instance, it is not a risk: it is a fact, a reality in several projects I have been involved.

The risk is to have many strikes in the construction site – the effect would be delays (impacting the “to be on time” objective) and payment of damages (impacting the “to be on budget” objective).

There is a proper way to fill the risk register: the use of risk metalanguage.

Risk should be expressed in sentences with this structure:

“Because of <cause>, <risk> might occur, which would lead to <effect>.

This language construction help bringing clarity and logic to the risk register. The previous example would become for instance:

“Because of <the presence of strong Workers Unions in the area>, <many strikes> might occur, which would lead to <delays and payment of damages>.”

There are some interesting implications. For instance this structure should be used in different languages – I would be curious to see if it can be replicated easily in all of them.

Additionally, it force you to think at the logical correlation between the causes of risks.

For instance, you can think at different situations were the same risk can have 2 different, independent causes:

“Because of <cause #1 AND cause #2>, <risk> might occur, which would lead to <effect>.”

I understand that in this case it should be split in 2 different entries of the Risk Register.

However in some situations two or more different causes need to be there at the same time for the risk to materialize:

“Because of <cause #1 PLUS cause #2>, <risk> might occur, which would lead to <effect>.”

Obviously you can increase the complexity ad libitum, as the same risk can have more than one effect:

“Because of <cause>, <risk> might occur, which would lead to <effect #1 AND effect #2>.”

I guess that also in this case the good practice is to split the sentence in two different entries for the Risk Register.

Rock slingers for a quicker trench sanding & backfill

This morning I found by chance this very interesting website (well, it is interesting if you like wind farm constructions…).

Basically it is an Australian company using “rock slingers” (that is, conveyors belts connected to a dumper) to backfill trenches mounted on small vehicles (2.5 meters wide). The equipment is made by CAS, an American company specialized in this kind of equipment.

It is a remotely controlled machine that can create the sandbed inside the trench accurately and at a great speed. According to the figures provided in the website the slinger can create 16 Km of bedding in a day, using up to 1000 tonnes of material.

I guess that they call it "slinger" because it can throw material at a quite remarkable distance (over 40 meters). Used in combination with one or two trencher it looks like it can lead to relevant savings, less labor and a more homogeneous distribution of the material.

Risk & contingencies - a brief introduction

One of the hot topics frequently discussed in the creation of the budget for big projects is the appropriate contingency level and how to estimate it.

My colleague Giuseppe had found an interesting paper online that has been the starting point for this post.

So, what are contingencies?

Some far reaching definitions consider different typologies of contingencies:

  1. Money in the budget
  2. Float in the schedule
  3. Tolerance in the technical specifications
  4. Tolerance in the quality
  5. Tolerance in the scope of work

Possibly this is a very broad approach, so I will focus only on the first point.

Monetary contingencies are added to the estimate “to allow for items and events for which the state, occurrence or effect is uncertain” (the definition is proposed by the Association for the Advancement of Cost Engineering).

Several concepts are usually excluded from the contingency budget:

  • Major changes in scope
  • Extraordinary events (such as the one indicated in the Force Majeure clauses)
  • Currency exchange risk (this is usually hedged or it is included in a different section of the budget)

Basically in the definition of contingencies the focus is on the “negative risks” that can create a loss if they materialize (“positive risks” is a fancy way to call the opportunities).

Identified vs unidentified risks

A key distinction should be made between identified risks and unidentified risks.

Identified risks are “known unknowns” – that is, risks you are aware of (a classic example would be the geotechnical risk.

Unidentified risks are more similar to “unknown unknowns” - risks that come from situations that are so unexpected and difficult to foreseen even to subject matter experts that they would not be considered.

On top of this type of risk, there are also risk that emerge later in time – as a result of our actions and decisions or as a result of actions and decisions of external agents.

For this reason risk management processes include periodical risk reviews: you do your best to identify all risk at the beginning of the project but the situation can evolve in unexpected ways.

Identified risks can (and should) be managed: they should be included in a risk register, with a quantification, a predefined plan if the risk materialize, an owner, etc.

Strategies for identified risks

Additionally, several strategies are possible for identified risk: they can be

Avoided (if a subcontractor has a poor financial status it can be removed from the bidders list)

Transferred (if the failure of the main transformer can put at risk the viability of the investment, a business continuity insurance can be purchased)

Shared (if a project is very big, possibly a Joint Venture could be a good choice)

Mitigated (if a construction technology is very complex it could be a poor choice in an emerging country, where an easier solution could be a less risky choice)

Accepted (in this case, usually a monetary reserve is created for the accepted risks).

What about unidentified risks?

Unidentified (“unknown”) risks are conceptually different. Even if a great effort has been made to identify all possible risks the experience show that when the project is finally built several unforeseen events will happen, impacting the budget.

The quantification of the contingency for these unknown risk is an hot topic.

 

Do not trust me: reliance of data in EPCs

EPC contracts frequently include a clause on the reliance of data. It has several formulations, but it usually looks something like this:

“Employer-Provided Information has been made available for reference only.

The Employer makes no warranty as to the accuracy, completeness and reliability of any information, data, statement in the Employer-Provided Information.”

The objective is clear – avoiding claims during construction if the data is wrong. In the most extreme cases even information critical to price a project (like for instance a geotechnical survey) should be considered as a “not rely upon” data.

How did we get to this point?

To give some context, the theory is that standard EPC contracts like the Silver FIDIC explicitly request to the subcontractor to study the Employer’s Requirements to find errors and omissions.

This is usually written in this way (you can find an example with the full text for instance in chapter 5 of the standard Silver FIDIC contact):

"The Contractor shall be deemed to have scrutinised, prior to the Base Date, the Employer’s Requirements (including design criteria and calculations, if any)."

But there are exceptions - in an effort to create a reasonable contract, although not as balances as the FIDIC Red or Yellow, the authors of the clause add:

"However, the Employer shall be responsible for the correctness of the following portions of the Employer’s Requirements and of the following data and information provided by (or on behalf of) the Employer:

(a) portions, data and information which are stated in the Contract as being immutable or the responsibility of the Employer;

(...)

(d) portions, data and information which cannot be verified by the Contractor, except as otherwise stated in the Contract."

In an effort to unload risks and responsibilities, Employers try to avoid being accountable for ALL information provided during the tender. Basically, the bidder cannot trust the tender documentation and should double check.

Unfortunately, in many situations bidders cannot verify independently the information provided. For instance to confirm the results of a geotechnical survey for a wind farm a bidder would have to invest thousands of dollars and one or two months of time to make a new set of boreholes and trial pits.

This is clearly unrealistic and unreasonable. Such clause, if stretched to the extreme, can have as a result extremely high prices (as the bidder will have to foresee the worst case scenario) or few bidders (as they will simply decline to bid).

I also suspect that in some situation the Employer decide not to circulate available information, probably following some twisted logic.

All in all I strongly believe that it would be in the interest of the Employer to avoid using the “not to be relied upon” clause on information like wind data, geotechnical survey, grid connection info, topographical survey, etc.

BoP strikes back: the increasing relevance of Balance of Plant

A key difference between combined cycle plants and wind or solar plants is the CAPEX / OPEX distribution. According on recent data of the American Department of Energy, for a combined cycle plant the CAPEX will be only around 25%, being the overwhelming majority of the investment in operational costs (that is, fuel) and maintenance.

The picture is different for wind farms and photovoltaic plants, were the fuel is free (still no taxes on wind and sun) and the majority of investment is needed upfront, with over 80% of CAPEX.

An interesting trend I am observing is the shift in the weight of Balance of Plant (BoP, more usually called Balance of Systems in the PV industry).

It is well known that the costs of photovoltaic modules and turbines are following a downwards trend. I do not see the same trend for BoP, with costs per MW decreasing at a slower pace.

In the figure above, you can see how the BoP share can be more than half of the CAPEX for rooftop solar. The numbers are coming from the Fraunhofer Institute, which include in BoP also land acquisition costs, permitting and legal cost, taxes, etc. Even if I disagree with this “broad” definition of BoP, the result is unchanged - the relative weigh of BoP in renewable is increasing.

What are the consequences?

In my opinion, the most relevant is that now BoP can really make of kill a deal. When it was approximately the 20% of the CAPEX even a big movement in the BoP budget was not really moving the numbers that much. However now the relative weight increased to over 40% in some project, and an expensive BoP can make a project economically inviable.

Is there something that we can do about it?

For several items probably not. For instance, wind farms on top of mountains will need expensive access roads, complicate earthworks with rock blasting, etc. While there is probably still some room to decrease the price of wind turbines (e.g. with a better supply chain) I do not see why civil works should be cheaper in the future.

The same apply to some electrical works items such as medium voltage cables, which are basically a commodity linked to the price of aluminium, steel, etc.

In conclusion, I think that in the next years we will see an increasing effort in engineering and optimization to lower the cost of BoP and insure the economic sustainability of projects.

Automatic cost estimator - the Holy Grail of BoP

Yesterday I had the pleasure to drink an overpriced coffe (2,90€ for an Espresso? Really?) with my good friend José Ramón. He told me that I’m not writing on the blog often enough so I’ve decided to make an effort and find some time to write this post.

The subject I have selected is an evergreen topic, the Holy Grail of BoP – the possibility to create a tool that could calculate quickly the cost of the BoP of a wind farm.

There is already a good amount of material on the subject online, for instance this website of the University of Strathclyde (Glasgow) that present a model created in collaboration with SgurrEnergy (now part of the Wood Group's Clean Energy business).

You can download the tool from their web or from this link for your convenience: BoP estimator tool

I have decided to take it as starting point to show why the task is not so easy and probably me and the other engineers in the team will not be substituted by an Excel file anytime soon.

The ultimate purpose of such models is to pick a small number of input (to make the tool usable) without losing to much in accuracy. The guys at the University decided to go for an extreme simplification and selected only 6 inputs:

  1. Number of WTGs
  2. Turbines Rating
  3. Km of new roads
  4. Km of existing roads
  5. Km of cabling to substation
  6. Km if cabling to grid

That is a very, very extreme oversimplification.

For instance, the model doesn’t keep into account the topography of the area (flat, hilly, mountainous) or other relevant factors (poor soils, inundable areas, etc.) and link the cost only to the rated power of the turbine. As a consequence the calculation of the crane pads cost show a big dispersion in prices (from 5.000 to 42.000 Pounds) and a very low R Squared value (0.26 – that is, the model isn’t explaining the correlation).

Additionally, the model doesn’t consider any monetary input but the output is monetary. I believe it’s rather hard to accept this simplification: for instance, around 50% of the price of cables is in raw materials like copper, that have a high volatility. This could easily bring a multi million inaccuracy.

Also, there is no such a thing as a standard substation – and this is why we have very good electrical engineer in the team. Even without considering the peculiarities of the local grids (something hard to ignore when they are weak, like in Australia) different customers have also different needs. A customer interested in business certainty will ask for redundancy in the substation – 2 main transformer instead of one, emergency “cold” spare transformer, etc.

Same for the foundations: there are currently so many technical solution in the market (precast, with rock anchors, braced, P&H, etc.) that it is really hard to find a correlation between wind turbine MW and foundation cost. There is so much money in foundation and so much pressure on prices that project specific foundations nowadays are the norm, not the exception.

Circular economy: use of wind turbines blades as combustible and mix material for cement production

One of the future challenges of wind energy is to find a solution to recycle old blades from decommissioned wind turbines. In this post I will try to summarize several possible alternatives and to describe in detail what I think is currently the best option: to use them as a component for the production of cement.

The development of this interesting technical solution started in 2005 when one of the largest wind turbine manufacturer asked LafargeHolcim (a global cement producer) to use the decommissioned blades in cement production plants.

In 2008 Geocycle (the business unit of LafargeHolcim specialized in use of waste to produce cement) launched the full scale development of the solution in the Lägerdorf cement plant, in Northern Germany.

The recycling start in the wind farm, where the blades are cut in 10 meters long pieces using a mobile cutting technique that reduce the generation of fine dust humidifying the area. The resulting water is collected and brought to the recycling plant together with the other materials.

The pieces are then transported by train or truck to a pre-processing plant.

Here the blades are cut before in segments with a length around 1 meter and subsequently shredded in smaller pieces, with a final length of some millimetres.

Metals, both ferrous and not ferrous, are separated automatically from the material flow by magnet and eddy current magnet devices.

Finally, the crushed blade dust is mixed with a humid substrate material made of other residues such as plastic labels and miscellaneous packaging materials. The purpose of this substrate is to homogenize and bind together the blade dust. The material composing the substrate are wastes from other processes such as paper recycling.

All process steps in the pre-treatment plant are fully automated and performed in a controlled atmosphere. This guarantee maximum levels of occupational health and safety.

The end product is put in the cement plant.

In the pre-calciner (an element of the cement plant that optimize fuel consumption) the resin matrix is used as an alternative fuel, substituting coal, petroleum coke, heavy fuel oil or natural gas.

At a temperature around 900 °C the resin burn transforming the blade fibers into ashes. These ashes are then going together with the rest of the calcinated raw materials - usually a mixture of limestone and clay - in the sintering zone of the cement kiln to produce the clinker.

Finally the clinker is ground into a fine powder together with a small amount of gypsum creating the final product, cement.

The cement produced using wind turbine blades is indistinguishable in terms of quality from the standard product obtained without using the blades. It can be sold in the market or used again in wind farms, for instance in wind turbines foundations, closing the loop.

This solution provide alternative raw materials for cement production, reducing the need for quarrying, stone crushing and transportation. The ash of a wind turbine blade consists mainly of silica (SiO2) and calcium oxide (CaO) and due to this substantial amounts of natural resources like carbonate rock (limestone) and clay (usually in the form of sand) can be saved.

Additionally, it has to be noted that this solution contribute to savings of fossil fuel: the resin of the blade provide energy (around 12 MJ/Kg). This value is approximately the half of hard coal:  therefore each tonne of rotor blade substitute around half a tonne of hard coal.

It’s important to note that this solution has been indicated in 2011 as the best option available in a Position Paper subscribed by the European Composites Industry Association and other relevant category associations.

The test phase was successfully concluded in 2009 and the solution has been made commercially available since 2010.

I see several interesting benefits associated with this technique:

  1. It is capable of dealing with large amounts of materials, since the cement plants can handle thousands of tons of material.
  2. It address as much as possible geographically local opportunities, because it’s operated by LafargeHolcim, a multinational company active worldwide with over 2000 operation sites.
  3. It is feasible right now, because it is a commercially mature solution currently used in Germany and neighbouring countries. Hundreds of blades have already been recycled.
  4. It is cost-effective, because is the second cheapest alternative after landfilling.
  5. It address the totality of the materials of the blades, because the resin matrix is used as combustible, the fibres become part of the cement and other elements (e.g. metallic components) are separated upfront in the pre-processing unit.
  6. It follow strict environmental and safety standards. This solution has been developed in Germany (one of the first country to ban the landfill of wind turbines blades) and it has been certified by third party bodies.

The final cost is function of several parameters, such as the number of blades to be recycled, the distance of the wind farm from the nearest cement plant and the need to adapt the pre-processing facility.

Therefore, final cost will depend on scale factors, logistic costs and pre-treatment costs.

There are several other alternative. I will try to summarize them in the next paragraphs.

Landfill: Blades are sectioned into pieces of suitable dimensions and placed in a landfill. Transportation and preliminary shredding is needed.

Here no material recovery is possible and some national and local legislation impose high landfill taxes or complete bans on composites material with high organic percentage such as the blades.

Additionally, legislation worldwide is expected to become stricter acting as a driver for the development of alternative solutions and the creation of a recycling market, either with an extended ban or with a sharp price increase for landfill of blades.

Mechanical processes: Blades are reduced in size and separate into powder and fibrous fractions via cutting, crushing, shredding, grinding, or milling processes.

Finer pieces are sorted and used as fillers or reinforcements or as fuel for thermal waste processes.

This is currently done on a limited commercial scale and theoretically it has a wide range of potential applications for material recovery.

However, from an economical point of view it is competing with the lower market cost of alternative virgin fillers (calcium carbonate, silica).

Additionally, it has been observed a reduction of the mechanical properties of the resulting material (lower stiffness and strength). For instance, concrete made using shredded blades instead of crushed stones has unsatisfactory mechanical properties.

The only viable mechanical process currently operating in the market, presented as the competitor solution, is the one offered by Global Fiberglass Solutions, detailed in the following paragraph.

Mechanical processes - Global Fiberglass Solutions (GFS) pellets: The blades are transformed into small pellets that are sold under the brand name of “EcoPoly Pellets”. They are usable in injection mold and extrusion manufacturing processes and they are made from a customized blend of wind turbine blade materials.

EcoPoly Pellets are made to order for final customers based on the requirements of the customer’s own manufacturing process. GFS has initiated distribution options for pellet and has already found several wind farmers owner willing to provide decommissioned blades as input material for pellet production.

I believe that this technique it is currently available only in the USA.

Thermal processes - incineration: Blade sections are incinerated at high temperatures – usually over 800 °C.

Organic substances are combusted and converted into non-combustible material (ash), flue gas and energy. This process is usually combined with energy production and heat recovery.

The ash could also be used as a substitute for aggregate in other applications recovering the material or landfilled. However, no economically viable uses have been found for the ashes (beside the proposed use as input for cement production).

The advantage is that there are already numerous incineration plants available and that it can be done at attractive prices.

The disadvantage is that no large parts can be incinerated, making necessary a preliminary shredding.

Additionally, blades have a high ash content (around 50% to 65%) that needs to be landfilled afterwards with the associated transport and landfill tax costs.

Finally, burning blades create residues that can cause problems in the gas filtering systems.

For all these reasons this solution has been discarded.

Thermal processes – Pyrolysis: Blades are sectioned into suitable dimensions and decomposed using heating ovens in an inert atmosphere (450-700 °C).

Material is recovered in the form of fibres which can be reused in other industries. For instance, in pilot tests they have been used for glues, paints and concrete. Other products of this process include syngas (that can later be combusted for electricity and heat recovery) and char (recycled as fertilizer)

Under certain boundary conditions the resin matrix can be transformed in an oil “Pyrolysis Oil”.

The problem with this technique is that significant amount of energy are needed to activate the pyrolysis, impacting the overall environmental value of the solution.

Furthermore, available laboratory tests show a degradation of the properties of the glass fibers and no secondary market has been found.

Additionally, test have been performed only at laboratory scale so it is considered a solution with a low technological maturity.

Thermochemical processes – Solvolysis: Chemical solvents (water, alcohol, acid) are used to break the resin matrix bonds at elevated temperatures (300-650 °C) and pressures conditions.

Among all the solvents currently tested, water appears as the most commonly used. In this case the process is called hydrolysis.

Fibre materials recovered have a similar strength and could be reused in other applications. The resins as well can be separated and combusted for energy recovery

As far as I am aware test have been performed only at laboratory scale, mainly with focus on carbon fibres composites.

Moreover, due to the low market price of glass fiber, no secondary market is existing.

Lightning protection of wind turbine blades

Image copyright of Elsevier & X.Bian. Published on “Numerical analysis of lightning attachment to wind turbine blade”

I have received a question from a reader regarding blades damaged by lightning.

Specifically, the blade has been damaged before commissioning.

At first sight the consequences could be less significant than usual, above all if the main crane is still on site and there is a set of spare blades available as happens frequently in big wind farms.

I would also guess that the turbine supplier will have to absorb the cost unless the risk was already transferred to the customer.

I would also like to elaborate a bit more on the topic as lightning is a frequent cause of damage to wind turbines (specifically to the blades, as they are hit in around 75% of the cases).

Lightning are created by the electric field between the bottom of the clouds (negative) and the ground (positive).

The potential difference is significant (some MV). However, due to the distance, the average electric field is weak.

As the electrical charges at the bottom of the clouds accumulate a “downward leader” (a channel permitting the flow of negative charges) start moving toward the ground.

If this stepped leader is somewhere nearby a wind turbine (or another similar structure) the second phase of the phenomenon may start: an “upward leader” from the blade connecting with the downward leader and closing the circuit.

This is the instant where the “return stroke” start and you usually see the majority of the light.

Coming back to the blades the standard technical solution consist in embedding a copper receptor connected to the ground at approximately 1 meter to the tip. This receptor has a diameter of at least 50 mm (pretty much like the copper cable used in the earthing of the foundations).

The problem is that this solution doesn't works always: sometimes the lightning hit another point of the blade. Even if the surface of the blade is supposedly non conductive it has been observed that, due to the presence of pollution and water, it can behave like a conductor.

RUTE precast modular wind turbine foundation

Some days ago I have been contacted by Doug Krause, founder of RUTE - a green start up proposing an interesting solution for the wind turbine foundation.

Taking inspiration from the technology used in bridge construction RUTE is proposing a system of post-tensioned beams connected to a central hub. Each beam has an anchor system connecting it to the soil, and the foundation is delivered to the wind farms in around 20 elements.

Among the benefit the fact that the system is modular, less prone to quality problems (it is manufactured in specialized plants) and, at least in principle, reusable after 20 to 30 years for a new foundation: the lifetime of the components is over 40 years.

Decommissioning is also probably easier with such structure, at least compared with a standard shallow foundation.

Installation times can be cut as well – as it is delivered hardened it is ready for installation in few days from the start of the works.

I had a look at the technical specifications and I have seen that the bottom of the excavation is at the same depth of the standard solution, so no savings here. I have also noticed that in some situation soil substitution could be needed.

I have seen in their website that there is already a full-scale prototype built, so it’s much more than a concept. It has been installed at the Palmers Creek Wind Farm (Minnesota) on a 2.5 MW GE turbine with a 90 meter hub height.

Addenda (10 June 2019): I've received an email from the founder of the company. I post it here for the benefit of all readers.

Thank you Francesco for noticing RUTE.
That's a picture of our rock anchor, bulb T girder, model TG. The one we built in Minnesota is a box girder style, BX Foundation. It behaves just like an inverted T, spread foundation.

RUTE's biggest value to the BOP contractor is time. So most of the foundation works can happen off the project books and schedule. So a project can close finance and be erecting towers the same month. We'll hope to prove that claim this year.

Apart from the main BOP driver, the facility owner can run a pro forma out 30 years, or 40 years, the normal term of the land lease. And in those cases a foundation with bridge design, like ours, lasts well past 40 years. That's just a function of the post-tensioning which keeps the concrete in permanent compression. So there's an order of magnitude less fatigue damage than conventional reinforced concrete.

I can share some pictures from inside the foundation. You can walk around inside it and inspect.

Best Regards,
Doug, RUTE

Wind farm optimization algorithms

I have always been amazed by the number of published papers, master thesis and documents focusing on the use of algorithms to optimize the layout of a wind farm. Some of them were proposed more than 25 years ago, showing a continuous, sustained interest in the topic.

I guess that the reason for such abundance is the stimulating difficulty of the problem and the fact that there are huge investments behind a wind farm.

From a mathematical perspective the problem is complex due to the type of variables involved, both discrete (you can have 30 or 31 turbine but not 30.5) and continuous (for instance, the length of cables). Additionally there are strong links between variables (for instance higher turbines = higher tower and foundation cost) so finding the “sweet spot” that maximize earnings is not a simple task.

Generally speaking, these algorithm try to maximize the profitability of the investment, usually expressed in terms of Net Present Value (NPV). Basically they compare the value of all expenditures during the life of the project “in today money” with all the earning “in today money” using a certain discount rate for cash flows in the future.

Expenses belong to two categories, capital expenses (CAPEX) and operational expenses (OPEX), while net earnings are function of the amount of power produced, the price of electricity and the electrical losses.

Therefore even a simplified model should try to minimize these expenses:

  • Wind turbine
    • Model (power curve)
    • Tower
    • Installation
  • Civil works
    • Foundations
    • Roads
  • Electrical works
    • MV cables
    • Substation
  • Operation & Maintenance

While maximizing the production, a mainly a function of:

  • Wind
  • Wind shear (of the speed of the wind increase with height)
  • Wake effect (how turbine interact with each other creating turbulences)

The interaction between all these variables is what makes the problem interesting.

To give a few examples,

  1. Packing the turbines densely in a small area will lower the cost of roads and cables but will create huge production losses due to the turbulences inducted by the turbines upwind.
  2. Using a higher tower should increase the production – unless the wind shear is low, in which case the additional tower and installation costs would off weight the benefits
  3. A certain position could be extremely productive – but it could be very far away from the substation (increasing the electrical losses ) or on the top of a steep hill (increasing the earthworks cost)

Additionally you have to decide the level of complexity of the model. For instance the foundation cost can be considered as:

  • A lump sum, equal for all turbine models. Under such assumption, you would see a benefit decreasing the number of turbines but not switching to a different WTG model.
  • A function if the wind turbine model (greater loads = greater foundation).
  • A function of wind turbine model, geotechnical parameters of the soil and unit cost of concrete of still. This latter option, although more precise, would probably make the model very difficult to handle.

I believe that a reasonable compromise between complexity of the model and quality of the result can be achieved using nested algorithms as proposed by these researchers.

In the first steps, only the variables related to the turbines (power curve, wind resource, availability and cost) are considered. Once the turbine model and the layout are fixed the civil and electrical works can be considered, defining the optimum position of the substation (to minimize cable length) and the shortest roads connecting the wind turbines.