Readers will be aware of the opposition to large hydro developments from ecologists and the debate that has opened in the environmental community. We have a trade-off between the environmental, social and economic benefits of a new renewable energy installation, versus the environmental and cultural impacts of the structure. The ‘best’ answer at any one time would be one reached by a healthy democratic process where opinions on each side are heard and evidence assessed equally. The flaws in that solution can then inform the next, and so on.
A case in point would be the recent scheme at Glendoe in Scotland, where First Minister Alex Salmond closed the dam gate in September 2008. Glendoe will play a crucial role in meeting Scotland’s and the UK’s renewable energy targets, but Energy Minister Jim Mather warned that ‘we are unlikely to see much in the way of further large scale developments.’
This is not to say large scale opportunities don’t exist, but that we may need to lay off the valley-flooding for a bit. In five or ten years it may be time to look again.
This puts the spotlight on small hydro (up to 5MW or so) where small impoundments or even run-of-river schemes will suffice. Engineering consultancies across the world are dusting off their ultrasonic transducers, but it seems many small landowners are put off by the upfront cost of a pre-feasibility study: £500 (US$765) to perhaps be told there’s no potential at all. And the poor old developers are having to visit ten unsuitable sites for every good one.
Then there’s the planning process, costs of development and opposition from fisheries which, in Scotland, mean all hydro is presumed guilty until proven innocent. This is not the route to Jim Mather’s ‘sustainable and profitable future in smaller and micro hydro schemes’.
Enter Hydrobot, a new approach to hydro site identification which could solve many of these problems. Hydrobot is a combined geographical information system (GIS) and financial assessment tool, designed to mimic a quarter of a million hydro engineers tramping up and down the hills of Scotland finding hydro sites. The model first hit the headlines when it was commissioned by the Forum for Renewable Energy Development in Scotland (FREDS), on behalf of the Scottish Government, to assess the nation’s remaining hydro potential. The resulting Scottish Hydropower Resource Study had the remit of identifying the main barriers to hydro development in Scotland [see p12]. Now that the government has verified hydro’s contribution to renewables targets could be significant, it is committing resources to tackling the barriers to hydro development.
Confidence and further competition within the industry should bring prices down, but developers may still waste valuable time and resources poring over large estates. Hydrobot can be applied to any land area within Scotland and the top sites supplied to those developers.
Developing Hydrobot
Hydrobot was first conceived as a university project, and operated as a series of processes rather than a single model. It was initially used to analyse the catchments of the North and South Esk near Edinburgh. After this, the model followed two very different development paths. In 2007 I qualified for the Starter for 6 initiative, which is run by the National Endowment for Science, Technology and the Arts (NESTA). Starter for 6 is an enterprise support project, and provided funding that allowed me to establish a web portal for Hydrobot, and to develop the model for micro hydro applications.
When FREDS launched a competition in late 2007 to survey Scotland for hydro potential, Hydrobot was an obvious contender as it was already running, albeit for micro hydro. Developing the model to cover the whole of Scotland and bolting on other required features was relatively fast and economic, compared to what it would have taken any other consortium to do the same job. Costing and calibration was directed by black-veatch, with environmental and grid impacts analysed by SISTech Ltd.
Needless to say, other development was put on hold during the undertaking of the Scottish Hydropower Resource Study. As a result the online version only went live on 29 October 2008. It will primarily identify micro hydro schemes to begin with, though the other features used in the Scottish study that model larger schemes will be phased in over time.
Meanwhile the core Hydrobot model is being used and developed for clients requiring a larger or more specific service. The basic service will cost £35 (US$54) for a 1km2 tile, including VAT. The client receives a pdf report including a map illustrating the layout of the scheme, and a table summarising the size, costs and profits for each identified scheme. There could be several schemes within that area. By comparison, a traditional hydro pre-feasibility study, including site visit, would cost between £3-600 (US$459-918).
How to build an artificial engineer
So how does it work? Obviously the exact algorithms are commercially sensitive, but here’s an outline. Hydrobot is based on a surface flow model derived from elevation data in 10m x 10m squares across the whole of Scotland. Every watercourse has been modelled to give 21 exceedence levels on the annual flow duration curve at any point. The accuracy of the predicted flows has been tested against measured flows away from established gauging stations, and also examined by the Scottish Environmental Protection Agency (SEPA).
A range of grid connections are possible, depending on the location and size of the scheme:
• Domestic connection at 240V or 400V.
• Connection to an existing 11kV line.
• Connection to the 33kV network by installing a new 11kV or 33kV line, and connecting either at an existing substation or by constructing a new substation.
The model takes two approaches to site identification. There are several hundred weir locations already listed by the model, and each of these is tested. Turbine size and hence power output are based on head and flow, but adjusted to take into account environmental sensitivity of the area. The costs of each element of the equipment, installation, connection and management costs are then calculated taking into account site conditions. These costs equations were empirically derived using data provided by Black & Veatch. The power output of the scheme is calculated, using one of four turbine efficiency curves depending on the head to flow ratio. This leads to the calculation of the revenue and, by applying a discount rate to future revenues and costs, to the net present value.
More importantly, Hydrobot identifies sites where there is no existing infrastructure. Firstly areas are selected with a minimum slope reflecting that of operational hydro schemes, based on experience and historical data. Hydrobot places a turbine at the lower extreme of the slope, and simulates a 20m penstock. The full cost and revenue calculations are conducted as for weir sites above, to produce a valuation of that layout. Hydrobot then extends the penstock to 40m and repeats the exercise, typically with a lower flow but a higher head. In this way the model simulates a range of penstocks up to 1.5km. If one or more profitable solutions have emerged, Hydrobot moves the turbine 20m upstream and repeats the whole exercise. And this process will continue, meaning that many hundreds of layouts may be tested within one area. When no more viable layouts can be found, all the solutions are compared to select the ‘best’, based on preferences that reflect the investment profiles of different client-groups.
Multiple intakes are simulated by attempting to join nearby tributaries into a scheme, re-evaluating the combination, and comparing to the sum of the parts. Other tricks involve removing existing schemes, and simulating off-grid schemes, though the latter was not used in the Scottish study.
But what about dams? Storage schemes have not been forgotten. For every suitable site, the surrounding terrain is examined to determine whether it would be suitable for construction of a dam. This is based on the profile of valleys in Scotland where dams have been constructed for hydro projects, and depends on the slope and height of the banks, the width of the valley floor, and the flow in the watercourse. Where a site is deemed suitable, the additional head and greater efficiency of the system are taken into account, as well as the additional costs, and the site is re-evaluated. Again, it is compared to run-of-river options at the same site to determine whether a dam would be preferable.
By modelling multiple intakes on run-of-river schemes, and with the inclusion of storage schemes, Hydrobot identified over 1000 potential small hydro schemes across Scotland of up to 5MW, with a handful of schemes larger than this.
Who gets what?
The approach that Hydrobot takes, as described above, models how an engineer might identify a micro or small hydro scheme. Therefore it is appropriate for four principle audiences:
• The small landowner or farmer, who may have one site in mind.
• The developer, who may want the top sites in their region.
• The larger estate owner, who wants to know the potential on their land and where to begin.
• Governments and local authorities, who want to know the contribution that hydro can make to their targets, and how to release the bottlenecks.
Previous clients include the Scottish Government, the Forestry Commission, and multiple small estate and farm owners, and we are in discussions with several local authorities as well as developers and utilities. For each of these users Hydrobot can help reduce the time and finances required to identify sites with potential. For some it is equally important to see how the results change as market conditions fluctuate or grants become available. After the initial customisation to users’ needs, Hydrobot can be re-run with different input values at very little extra cost. This allows developers and landowners to quantify risk in more detail, which is especially relevant in the current economic climate.
A further benefit to re-running the model is the change in our other climate – the weather. As the flow regimes in river systems alter, becoming more extreme in their flood-drought cycles in the UK, the viability of many hydro schemes must be re-examined. Flow regimes can be adjusted for a whole country with Hydrobot, and the site identification process repeated. In this way, renewable energy policy can also be risk-reduced.
Future developments
Automated hydro site identification has taken a leap forward with the modelling of Scotland using Hydrobot in 2008. The resulting report champions the potential for grid-connected small hydro to contribute to national renewables targets. Hydrobot was developed to answer questions being asked in Scotland at this time, but what other questions are being asked?
• What about the potential for off-grid schemes?
• What about private networks and community schemes?
• Can you model other countries? Across the whole world?
• Can you model pumped storage?
• What about combining wind and pumped storage?
• Can you model the water treatment network?
• Don’t forget large hydro!
As a consultant, my answer to all of the above questions is ‘yes, of course’. And as the model becomes more sophisticated with each future version, it is likely that Scotland’s total resource will be refined. There may be an increase in the total as gaps such as off-grid and large hydro are filled. We would hope that the total will start to reduce as more and more of the identified sites are developed. We would hope to be producing results for other countries too, so that we start to put a clear number on the global potential for hydro.
But which question is being asked most urgently? This will determine the direction that Hydrobot takes over the next 12 months and beyond. I’m sorry, but I have to say it – Hydrobot will go with the flow.
Nick Forrest is Managing Director of Nick Forrest Associates Ltd and Hydrobot Ltd. He is also a director of babyHydro Ltd, a new full services company established to develop small scale hydro schemes in Scotland from feasibility to operation
The Hydrobot online service is available at http://www.nickforrestassoc.co.uk.
For further information please contact nick.forrest100@btinternet.com
