INOM
EXAMENSARBETE ENERGI OCH MILJÖ, AVANCERAD NIVÅ, 30 HP
STOCKHOLM SVERIGE 2017,
Frequency regulation and its effects on fish stranding
A hydrological modelling example of a Swedish river
JOHANNA EWERTZH
KTH
SKOLAN FÖR ARKITEKTUR OCH SAMHÄLLSBYGGNAD
Johanna Ewertzh
Master of Science Thesis
STOCKHOLM /2017/
Frequency regulation and its effects on fish stranding
A hydrological modelling example of a Swedish river
PRESENTED AT
INDUSTRIAL ECOLOGY
ROYAL INSTITUTE OF TECHNOLOGY
Supervisor:
Daniel Franzen Examiner:
Monika Olsson
2 TRITA-IM-EX 2017:20
Industrial Ecology,
Royal Institute of Technology www.ima.kth.se
3
Master Thesis project report
Title: Frequency regulation and its effects on fish stranding – A hydrological modelling example of a Swedish river
Public version
Author: Johanna Ewertzh
Granted by Fortum Sverige AB, for public use
_________________________________
Hans Bjerhag Date: 2017-08-16
4
Abstract
This is a master thesis aiming to answer how frequency regulation at a specific hydropower plant affect the risk of stranding for two common Swedish fish species: Trout and Grayling (Salmo trutta and Thymallus thymallus).
The study comprises a literature review and a hydrological modelling in primarily Hec-Ras. Frequency regulation is necessary in order to keep the electrical power system in balance. Frequency regulation in Sweden is handled to 99 % by hydropower. The energy company Fortum has several hydropower plants, distributed all over Sweden. In collaboration with Fortum a hydrological model was created for one of its plants and the downstream river reach. Reference scenarios for both summer and winter were then simulated and the result showed that the fish most likely were at risk of becoming stranded at certain times. The risk of stranding was more linked to the normal regulations taking place du to for example increased consumption of electricity in the river rather than isolated frequency control. An extreme scenario was also investigated and the water level change rate in that scenario was beyond the limits identified as what fish may withstand. When comparing the outcome of the simulations to the preferences of fish it becomes clear that mainly spawning and juvenile fish are at risk of becoming stranded.
5
Sammanfattning
Detta är ett examensarbete som undersöker hur frekvensreglering påverkar fiskarterna harr och öring (Thymallus thymallus och Salmo trutta) avseende strandningsrisken. Arbetet har gjorts som en kombinerad litteraturstudie och modellarbete i den hydrologiska programvaran Hec-Ras. Frekvensreglering är nödvändig för att hålla frekvensen i elkraftsystemet stabil. Frekvensen påverkas av produktion och konsumtion av elektricitet, en förutsättning för ett system i balans är att dessa är lika stora. Frekvensregleringen jämnar ut de skillnader som finns och detta görs i Sverige till 99 % med hjälp av vattenkraften. Energibolaget Fortum, som har vattenkraftverk utspridda över stora delar av Sverige, har varit samarbetspartner gällande detta
examensarbete och den studie som har gjorts har ägt rum vid ett av Fortums vattenkraftverk och dess nedströmsmiljö. En hydrologisk modell skapades för detta område och mätdata från både sommar och vinter nyttjades. Resultatet visade att risken för strandning finns under de båda årstiderna vid vissa tillfällen, dock ej som konsekvens av enbart frekvensreglering. Ett extremt scenario undersöktes också, i vilket
förändringshastigheten av vattenytan var större än vad som väntas vara bra för fisk dvs. större än 13 cm/h. När resultaten jämfördes med habitatpreferenser för harr och öring syntes det att lekande och ung fisk var de som var mest utsatta. Resultatet ska dock mer ses som vägledande än som en definitiv sanning då fel i modellen, brister i data och andra aspekter påverkar. Vidare studier är nödvändiga för att kunna dra direkta slutsatser.
Vidare studier skulle kunna vara om älvens fysiska attribut manipulerades och hur detta kan påverka risken för strandning.
6
Table of Contents
1.0 Introduction ... 10
1.1 Purpose of this study ... 10
1.2 Acknowledgements ... 10
2.0 Theoretical background ... 11
2.1 The Nordic electrical system ... 11
2.1.2 The Swedish power grid ... 11
2.2 Frequency control ... 11
2.2.1 Inertia ... 12
2.2.2 Primary frequency control ... 12
2.2.3 Secondary frequency control ... 12
2.2.4 Tertiary frequency control ... 12
2.2.5 The future of frequency control ... 13
2.3 Hydropower in Sweden ... 13
2.3.1 Comparison of electricity production ... 14
2.3.2 Fortum's hydropower ... 14
2.4 Environmental flows ... 14
2.5 Trout ... 15
2.5.1 Distribution ... 15
2.5.2 Diet ... 15
2.5.3 Reproduction ... 15
2.5.4 Habitat preferences and requirements ... 16
2.5.5 Restrictions and regulations ... 16
2.6 Grayling ... 16
2.6.1 Distribution ... 17
2.6.2 Diet ... 17
2.6.3 Reproduction ... 17
2.6.4 Habitat preferences and requirements ... 18
2.6.5 Restrictions and regulations ... 18
2.7 Habitat changes caused by hydropower ... 18
2.7.1 Water temperature ... 18
2.7.2 Dissolved oxygen levels ... 19
2.7.3 Hydropeaking ... 20
7
2.7.4 Stranding ... 20
2.7.5 Properties affecting stranding... 20
2.8 Environmental quality standards in Sweden ... 20
2.9 Ecological status and potential ... 21
2.9.1 Hydrological parameters ... 21
2.9.2 Water level change rate... 22
3.0 Aims and objectives ... 24
3.1 Specific research question ... 24
3.2 Aim ... 24
3.3 Objectives ... 24
3.4 Project plan ... 24
4.0 Method ... 25
4.1 Literature review ... 25
4.2 Interviews and correspondence ... 25
4.2.1 Field trip ... 25
4.3 Modelling ... 25
4.3.1 Software ... 25
4.3.2 ArcGIS ... 25
4.3.3 Hec-Ras ... 25
4.3.4 Data input ... 26
4.3.3 Assumptions ... 26
4.3.4 Calibration and validation ... 28
4.3.5 Model setup ... 28
4.4 Limitations of the study ... 29
4.4.1 Geographical limitations ... 29
4.4.2 Limited number of species ... 29
4.4.3 Limited data ... 29
4.5 Choice of location ... 29
4.6 Stakeholders ... 29
5.0 Results ... 31
5.1 Steady model ... 31
5.2 Unsteady model ... 36
5.2.1 Reference scenarios ... 37
8
5.2.2 January ... 37
5.2.3 June ... 39
5.3.4 Worst case scenario ... 41
5.3.5 Isolated frequency regulation ... 42
5.3 Model outcome linked to habitat preferences ... 44
5.4 Errors ... 45
6.0 Discussion ... 47
6.1 Discussion of results ... 47
6.1.2 Choice of reference scenarios ... 47
6.2 The behaviour of the frequency ... 48
6.2.1 A drop in frequency ... 48
6.2.2 Peaks in frequency ... 49
6.3 Data quality ... 49
6.3.1 Elevation/ terrain data... 49
6.3.2. Water level and flow data ... 50
6.3.3 Missing data ... 50
6.4 Further studies ... 50
6.5 Relevance of this study... 51
6.6 Other aspects affecting water levels and flows ... 51
6.6.1 Machine or grid failure ... 51
6.6.2 Upstream conditions ... 51
6.7 Reducing risk of stranding ... 52
6.8 Credibility of references ... 52
6.9 Water level change rate ... 52
7.0 Conclusion and recommendation ... 53
7.1 Conclusion ... 53
7.2 Recommendations ... 53
8.0 References ... 54
8.1 Written references ... 54
8.2 Picture references ... 56
Appendix 1 ... 57
A.1 Steps in constructing the model ... 57
9
10
1.0 Introduction
1.1 Purpose of this study
The electrical power system is undergoing large changes as renewable energy sources, such as solar and wind power, are introduced. The challenge with these new sources of electricity is that production happens
momentarily, when the sun shines or the wind blows. To have an energy system with low carbon dioxide emissions is desirable in order to limit the climate change and environmental impacts of our energy system, but the electrical power system cannot function when it’s not balanced. The balancing of the electrical power system in the Nordic and Baltic countries are to 99 % handled by hydro power. The hydro power has very low carbon dioxide emissions, much like the other renewable energy sources, but it has local impact on the environment (Svensk Energi, 2015).
In this thesis the connection between frequency control handled by hydro power and the stranding of fish has been investigated. The two investigated fish species will be Trout and Grayling (Salmon trutta and Thymallus thymallus). The study consists of a combined literature review and hydrological modelling in the software ArcGIS and Hec-Ras.
1.2 Acknowledgements
I would like to say a special thank you to my three supervisors; Daniel Franzén at The Royal Institute of Technology (KTH) in Stockholm, Hans Bjerhag at Fortum and Anders Johansson at SWECO. Your support has been most important to me.
Furthermore I would like to thank:
Fortum’s IT support – for always helping me with the software issues that I have faced during the work with this thesis
Helena Björkman – for helping me with the concept and design of the hydrological model
Kristina Berg – for your valuable help with ArcGIS
Rickard Eklund – for providing me with validation data
Per Andersson at Vattenregleringsföretagen – for providing me with validation data
The staff at the investigated hydro power plant – you made the field trip a highlight of this work
Magnus Jewert – for the delivery of the elevation data
Dag Cederborg – for your valuable input
My opponents – for helping me refining the report
Monika Olsson – for being the examiner of this report
Davide Attebrant Sbrzesny – for being the best moral support
Lastly I would like to direct a great thank you to Heini Auvinen, without you the model in this thesis would not exist. Your help has been invaluable. A thousand thanks for the time you have spent on me and my project.
11
2.0 Theoretical background
In this section some theory is presented which is necessary in order to understand the result of this study.
Firstly the electrical system and the concept of frequency control will be presented. It is then followed by a short background of Swedish hydro power before describing the two fish species included in this study; Trout and Grayling. How the hydropower affect the fish is then explained together with different hydrological concepts. Lastly, the Swedish environmental quality standards and ecological status will be explained.
2.1 The Nordic electrical system
The Swedish electrical power system and its grid is very much connected to and involved with the electrical power systems of the other Nordic countries (not including Iceland and some parts of Denmark). It therefore makes more sense to talk about the Nordic electrical power system rather than the Swedish electrical power system. On the first of January 1996 new rules about trading and producing electricity became official and the state monopolies that earlier had been limiting these things, were gone. A Nordic power exchange and trading platform was introduced: Nordpool. (Svensk energi, 2016a). Nordpool is owned by all the Nordic transmission system operators (TSO); Svenska Kraftnät, Fingrid Oy, Statnett SF and Energinet together with the Baltic transmission operators; Elering, Litgrid and Augustprieguma tikles. Today 380 companies from 20 different countries trade on the markets in Baltic and Nordic region (Nordpool, 2016).
2.1.2 The Swedish power grid
Even though it is hard to talk about a Swedish electrical power system, since the inputs and outputs to the system is distributed all over Sweden, Finland, Norway and large parts of Denmark, the geographical boundaries have been important for the grid system. Therefore, it is possible to differentiate the Swedish power system as a separate system when it comes to the grid of power lines. Every Nordic country has their own transmission operator, in Sweden that is Svenska Kraftnät (SvK). The rules and regulations considering the power grid did not develop towards an open market in the same pace as the production and trading of
electricity. Today there are several owners of the power grid in Sweden, the owner with the least amount of power lines has approximately 3 km whilst the owner with the most has 135 000 km of power lines. There are three levels of high voltage lines in Sweden,: local, regional and the transmission grid, adding up to the grid . The transmission grid is owned and operated by Svenska Kraftnät, whilst the other ones are owned and operated by 160 different companies (Svensk energi, 2016b).
2.2 Frequency control
Even though the Swedish grid has many owners, there is only one unit that is responsible for the balance of the grid and that is Svenska Kraftnät. The consumption and production of electricity must at all times be equal, otherwise it is reflected in the frequency. The ideal frequency of the Nordic grid is 50,0 Hz. It is desirable to keep the frequency around 50 Hz in order not to harm important components of the system. The allowed normal operations interval for the frequency is between 49,9 and 50,1 Hz. The accepted deviation at normal operation is in other words ± 0,1 Hz, and frequency control is actively used to keep the frequency within this span. When the frequency deviates further the frequency regulation is again used to get the frequency back on track, but with a more strict set of regulating parameters. Depending on how large the deviation is, different protective measures of frequency control can be used. There are three levels of frequency control in Sweden.
These are primary, secondary and tertiary control. The primary is the first one activated, followed by the secondary and at last the tertiary (Persic, 2007).
12 2.2.1 Inertia
The most important stabilizing factor in the Nordic system is the inertia. It can easily be explained as the resistance to change within the system. In more detail, the inertia in the electrical system is the inertia of the moving mechanical parts of the system such as turbines and generators. The inertia of the moving parts means that it requires energy to increase the rotating speed of turbines and generators and it releases energy to slow them down. To measure the inertia there is a constant of inertia H, which is measured in seconds. The constant is used to calculate how long it would take for the system to go from a nominal speed to entirely stationary.
The inertia is necessary for the electrical system since it acts as a short-time buffer for changes. The stored inertia in the electrical system is approximately 60-90 MWh (IVA, 2016).
2.2.2 Primary frequency control
The first frequency control to kick in is the primary one. It is basically always activated. It is most commonly known as FCR-N or FCR-D which stands for Frequency Containment Reserve for Normal operation and Frequency Containment Reserve for Disturbed operation, respectively. It becomes activated within seconds after a sudden load or consumption change occur. This frequency control also goes under the name automatic active reserve since it becomes activated automatically when the load changes. Primary frequency control is handled separately for every power plant that is equipped with the right technology and equipment. The power plant then senses the change in frequency and adjust its production after it. In reality this means that if the frequency drops the production will increase and if the frequency rise the production will decrease. The primary frequency control capacity is located mainly in Norway and Sweden. The total capacity within the Nordic grid should be at a minimum 6 000 MW/Hz and it is the TSOs’ task to plan that this amount is available in the system (Persic, 2007).
2.2.3 Secondary frequency control
The secondary frequency control is slower to kick in than the primary one. Rather than seconds this regulating power takes minutes before becoming active. The secondary frequency control is also known as FRR-A which stands for Frequency Restoration Reserve Automatic. It was introduced to the Nordic market in 2013. Before that the restoration of the frequency was handled manually (Fingrid, 2016). The amount of FRR-A frequency control that all Nordic countries are supposed to be able to deliver is 100 MW. Out of these 100 MW, Sweden is responsible to deliver 39 MW (Svenska Kraftnät, 2013). The most important purpose with this frequency control is to restore balance to the system in the long-run. The primary regulation can restore the balance to the grid, but when this is done a certain amount of the reserves are consumed and the nominal frequency is unbalanced. It is the secondary frequency control's purpose to make sure that the primary frequency control has the chance to restore itself (Persic, 2007).
2.2.4 Tertiary frequency control
The tertiary frequency control is mainly activated when there are large disturbances in the system. The tertiary frequency control is also known as FRR-M which stands for Frequency Restoration Reserve Manual, and as described in the name it is activated manually. It can require disconnection of certain energy-intense industries or starting up of standby power plants (Fingrid, 2016). The purpose of this frequency restoration is to restore or replace the automatic frequency restoration reserve and other frequency control so that it can be used again within ten minutes. The time to carry out the FRR-M is ten minutes, a fairly long time when talking about electricity that is produced and consumed momentarily. It was due to this long time gap between ordering the FRR-M and actually receiving the restoration that FRR-A was introduced (Byström, 2017). The tertiary
frequency control also aims to restore the time deviation. The time deviation is the accumulated deviation of
13
the nominal frequency and it shows if there has been imbalance in the system. It is the responsibility of the TSOs to make sure that the time deviation is not too large. The TSOs in Sweden, Denmark, Finland and Norway have an agreement that the time deviation should not be more than ± 30 seconds. If this were to be exceeded it can be viewed as an indicator that the average frequency is not 50 Hz (Persic, 2007).
2.2.5 The future of frequency control
As the energy industry moves towards a renewable future with increased amounts of wind and solar power in the system, the need for frequency control becomes more important. Wind and solar power cannot be
controlled in the same sense as hydro power. The production happens momentarily when the sun is shining or when the wind is blowing, although new battery technology connected to the solar power might have a vital effect for the grid in the future. Creating accurate prognosis of the weather is one why of gaining more control over the renewables, but the weather can be unreliable and the electrical grid must be reliable. The predictions for the future is that more frequency control will be needed and that it needs to work even faster than it does today. The price of electricity are set hourly in today’s market. This might change in the future in order not to change the load on the grid as rapidly when the prices go heavily up or down (Saarinen, 2017). Almost all frequency control is handled by hydropower. In 2013 the hydropower was responsible for 100 % of the automatic frequency control and 99 % of the manual frequency control. 255 hydropower plants spread out over the surface of Sweden with an installed output of 14 700 MW is standing for basically the entire frequency control of Sweden (Nyberg, 2014). The use of hydro power for frequency control is not predicted to decrease, on the opposite it is most likely to increase within the future. The transmission system operator SvK states that it is important that the EU framework is adapted, so that the future for hydropower as main frequency
controller remains (Svenska Kraftnät, 2015).
2.3 Hydropower in Sweden
Hydropower stands for approximately 50 % of the electricity energy production in the Nordic system and around 45 % of the production in Sweden. It is important to the energy system for several reasons. The first reason is the flexibility of the hydropower, which enables it to tackle changes in load and generation. The second reason is the low environmental impact in terms of carbon dioxide emissions, which is one of the reasons that it is possible for the Nordic countries to have an electric system with rather low contributions to climate change. A third reason is especially important, as other sectors such as transportation, moves towards an electrified system rather than a fossil fuel based system (Fortum, 2015).
Even though the environmental impacts from hydropower in Sweden are low in terms of greenhouse gas emissions, there are local environmental impacts that need to be considered when evaluating the total
environmental impact. The largest impact is the natural land and water transformation that takes place when a hydropower plant and reservoir is built. It changes the surroundings by flooding or drying when creating the dam and controlling the flow of the river. The impact from operating the hydropower plant is marginal, especially when comparing the initial damage (Malm Renöfält, 2009).
In order to minimize the environmental damage from hydropower, there are several laws and regulations.
Joining the European Union has also changed the laws and standards that the Swedish authorities work with, it added new pressure on how the rivers of the country are managed. In Sweden a total of 42 rivers or certain reaches of rivers are protected from hydropower expansion, however the majority of these rivers have already been altered by humans. Out of these 42 rivers and reaches four are mentioned as "national rivers" with similar status as national parks. These four rivers are Torneälven, Kalixälven, Piteälven and Vindeälven (Swedish EPA, 2016).
14 2.3.1 Comparison of electricity production
In order to gain a better understanding of the role of hydropower in Sweden one can study Table 1. The table describes how large the production, installed output, carbon dioxide emissions and expected availability are of the different types of electricity production sources available in Sweden. The numbers found in Table 1 are from the year of 2015 and can thus have changed slightly, however it is mainly the numbers considering wind and solar that might have changed due to further development. When studying Table 1 it becomes clear that the hydropower is responsible for a large share of the electricity production in Sweden and that the
environmental impact in terms of carbon dioxide emissions is among the lowest.
Table 1: Features of the Swedish energy mix, as of the 1/1 2015 (Svensk Energi, 2015).
Type of power
Hydro Nuclear CHP* Wind Condensing Solar
Swedish production
50-80 TWh/
year
50-70 TWh/
year
13-18 TWh/
year
14 TWh/
year
0,5-0,9 TWh/ year
0,1 TWh/
year Installed
output
16 155 MW 9 528 MW 5 056 MW 5 420 MW 3 311 MW 79 MW CO2
emissions
9 g/kWh 5 g/kwh 15 g/kwh 15 g/kwh 930-1 270 g/kwh
46 g/kwh Expected
availability at high demand
85 % 84 % 75 % 11 % 90 % 0 %
*Combined heat and power plant
2.3.2 Fortum's hydropower
Fortum is the second largest hydropower operator in Sweden. Only the state-owned company Vattenfall has more hydropower. 126 hydropower plants are owned or co-owned by Fortum and the total installed output is approximately 3 600 MW. On a yearly basis these 126 hydropower plants generate 15 TWh, out of these 13 TWh is Fortum’s share. The 2 TWh in difference goes to other shareholders in the co-owned hydropower plants (Bjerhag, 2017). The majority of Fortum’s hydropower plants are located in the rivers Klarälven, Dalälven, Indalsälven and Ljusnan. Fortum also owns hydropower in Finland, but not as much as in Sweden. Fortum is continuously working to increase the efficiency in its hydropower plants. Improved production will lead to a net increase in power of 100-150 MW until 2020 by renovation and increasing efficiency (Fortum, 2016).
2.4 Environmental flows
Humans have interacted with flows and rivers and altered their flows for more than 100 years. Historically the rivers have been utilised by mills, sawmills, forgeries and for transportation. On a global scale it is estimated that more than 50 % out of all accessible surface water is already appropriated by humanity and this number is estimated to increase within the near future. Out of the world's most primary watersheds 46 % had at least one dam of a large scale (Tharme, 2003). As mentioned earlier in this report, the situation of exploiting rivers and water flows to large extent is restricted in Sweden. 42 rivers are protected from hydropower expansion and four rivers have the classification "national rivers" (EPA, 2016).
The general concept of environmental flows is that enough water is let through the rivers to ensure all the benefits in the downstream area from environmental, social and economic perspectives. It is an
interdisciplinary approach that includes a number of disciplines such as; engineering, economy, ecology,
15
hydrology, law, political science and communication. It requires negotiations between all the stakeholders using the river or its water. The outcome of environmental flow is normally an altered management of the river with the ambition to vitalize the ecosystems and create optimal balance between all the users of the river. The environmental flow was mainly created as a compromise and to satisfy all stakeholders interested in the water.
In many cases around the water runs over country boarders and several cultures are dependent on the same river (IUCN, 2003). In Sweden that is not the case. Due to geographical conditions the only possible sharing partner in terms of countries is Norway and they do not lack water nor rivers. However, there are several stakeholders that have different interests in the rivers. These stakeholders will be listed and shortly discussed under section "4.6 Stakeholders" later in this report.
2.5 Trout
One of the two chosen fish species to further investigate in this study is Trout, Salmo trutta in Latin. In this section general information about Trout will be presented. The main focus will be on the habitat preferences since this is the aspect mainly affected by rapidly increased or decreased flow from primary frequency control.
However, to understand the habitat preferences, information about distribution, diet and reproduction will be further explained.
Figure 1: The appearance of a Trout (SAMWM, 2016a).
2.5.1 Distribution
The Trout can be found in all of Sweden. It can be found in the rivers flowing through the northern mountains to the most southernmost tip of Sweden. It can live in fresh, brackish and salt water and is thus found in both rivers, lakes, the Baltic sea and Kattegat. The recreational fishing is dominating the catch of Trout in Sweden.
The maximum age of a Trout has been measured to 18 years but seven to ten years is a more likely age. The fish can weigh up to 15 kg and a picture of the fish can be seen in Figure 1 (SLU, 2015).
2.5.2 Diet
All Trout is hatched in running water, streams or rivers. As young individuals, they feed on insects and other small food that goes with the water. As they grow they consume larger prey mainly consisting of other fish. If the Trout migrate to the sea the diet is mainly Herring and Sprat. If the fish do not migrate and stay in rivers or near-by lakes the diet is made up by Vendace, Stickleback and Smelt instead. Insects and benthos can also continue to be a substantial part of the diet for river Trout (SLU, 2015).
2.5.3 Reproduction
The Trout spawn in many different types of running water. The spawning takes place during the autumn and the fertilized fish eggs hatch during the spring. The fish become sexually mature at the age of two to five years.
16
It takes approximately one to five years for the Trout to start migrating to the sea or lakes, where it stays for about half a year to three years before it migrates back to the streams to spawn. Middle sized Trout can migrate at most 200 km (SLU, 2015). When the spawning takes place the female digs a small pit or pocket in the river bed with her fin and place her eggs there. An ideal river bed mainly consist of pebbles or gravel. The male then fertilize the eggs with his milt (Länsstyrelsen Västra Götaland, 2009).
2.5.4 Habitat preferences and requirements
The spawning Trout prefer water velocities between 0,2 and 1,1 meter/second. The velocity needs to be high enough to ensure that the eggs buried in the substrate get oxygenated, but not too high because then the eggs are at risk of becoming bared and then eaten. The substrate of the spawning Trout must be a mix of gravel and small pebbles for the same reason. Sand or clay does not allow oxygenated water to run through it to the same extent. It is also harder for the fish to move around with their fins. The preferred size of the substrate is
therefore 2 – 64 mm in diameter. The spawning Trout also tends to place their eggs in areas that are rather shallow. The depths most frequently occurring are 10 – 80 cm (Person, 2013).
The juvenile Trout is hatched at the same place as its egg has been buried. It thus starts its life at the same location as the spawning ground. During the first year the Trout prefers to be in locations with rather low water velocity. This is due to the fact that the juvenile Trout has limited swimming strength due to its size. The
preferred water velocity are velocities below 0,2 meter/ second. The juvenile Trout also tend to stay in rather shallow areas in order to avoid high water velocities and larger predators. Depths between 15 and 80 cm are to be preferred (Person, 2013).
The adult Trout do not have as specific preferences regarding habitats compared to the spawning and the juvenile Trout. They can be found in both rivers and lakes and seem to have no particular preference regarding substrate (Person, 2013).
2.5.5 Restrictions and regulations
The fishing of Trout is limited in several places in Sweden at different times. Fishing with gillnets, also known as phishing is only allowed in the four largest lakes in Sweden (Vänern, Vättern, Mälaren and Hjälmaren) to a depth of 100 meters and 180 meters in Storsjön located in Jämtland. Fishing for Trout is prohibited in Gullspångsälven all year round. It is also prohibited in Vänern, Vättern and Mälaren during the period 15th of September to 31st of December. In Storsjön and connected streams fishing for Trout is also prohibited from 1st of September to 31st of December. There are also certain areas near these lakes and streams where fishing is not allowed at all. There are also minimum dimensions of caught Trout in all these rivers so that the stock will survive. Many of these restrictions and regulations were introduced during the early 2000’s. A lot of them were originally introduced in order to protect the Char (Salvelinus alpinus) which is closely related to the Trout (SLU, 2015).
2.6 Grayling
The second fish that will be focused on in this study is the Grayling, named Thymallus thymallus in Latin after the smell that is similar to Thyme. This section contains general information about the grayling but has the habitat preferences as a main focus for the same reason as for Trout. Even if the Grayling to some extent share the same habitats as the Trout they do not have a similar appearance. The look of a classic Grayling can be seen in Figure 2 below.
17 Figure 2: The appearance of Grayling (SAMWM, 2016b) 2.6.1 Distribution
In Sweden one can find the fresh water fish Grayling all over the country, it is however more represented in the north. Even though it is a fresh water species it can be found in the brackish water of the Bothnian Sea. Very little is known about the Grayling living at sea compared to the documentations made about the Grayling living in rivers and lakes. The Grayling has a complex migration behaviour. During the summer the fish is stationary, remaining in areas with river beds mainly consisting of stones and gravel. As autumn comes, the fish migrate to their wintering area which typically is a calmer area. The wintering area can be located several kilometres from the summer area. As spring arrives the Grayling starts moving again to the spawning area. The normal
maximum age of the Grayling is 15 years, in rare cases individuals with the age of 20 years has been observed (Viklund, n.d.).
2.6.2 Diet
The Grayling is a sight feeder, which means that it uses its eye sight to catch prey. Newly hatched fish start eating pray approximately four days after hatching. They then mainly eat immature aquatic insects and fly larvae. As they grow they depend mainly on insects, both benthic and terrestrial, but have also been seen to consume fish eggs, small fish and small mammals. They eat and hunt all year round but as they depend on eye sight the hunting is less effective during the dark winter underneath ice and snow (U.S. fish and wildlife services, 1985).
2.6.3 Reproduction
The Grayling spawn annually in streams. The spawning starts with the male establishing a spawning territory that varies in size from 1 to 10 m2. The female then spread her eggs over the male's territory and he fertilizes them. As the egg are settling to the stream bottom, sand and small gravel is stirred up by the fish to create a light cover of the eggs. The female also presses her body towards the bottom when releasing the eggs in order to put them a few centimetres under the surface. The spawning period varies and has been observed to take 2 to 24 days. The preferred water temperature for spawning is 2-10 degrees Celsius, however more commonly taking place during temperatures of 6-10 degrees Celsius. At the age of 2-5 years the Grayling becomes sexually mature, males normally before females. Much like the Trout the Grayling want to have a river bed that consists mainly of pebbles and gravel, they are not as particular about this as the Trout and can also spawn over soft river beds with mud and vegetation (U.S. fish and wildlife services, 1985). The spawning takes place in very shallow waters, it is even common to see the dorsal fin pop out of the water when the Grayling is spawning.
The Grayling's spawning behaviour stands out compared to other salmonid fish because it takes place during spring instead of the autumn. When the ice starts to break and melt, then the Grayling migrate to the place where it once hatched (Viklund, n.d.).
18 2.6.4 Habitat preferences and requirements
The Grayling has different habitat preferences depending on the seasons and age. The spawning Grayling desires a mean velocity of the water of 0,4-0,7 m/s. The most dominant substrate should be of the size 16-32 mm and the water depth should be 10-110 cm. The strict preferences for velocity and substrate is due to two reasons: 1) the female needs to be able to bury her eggs in the fairly porous substrate and 2) the buried eggs need to be oxygenated (Nykänen, 2004).
The larval Grayling cannot withstand too high water velocities and prefer velocities below 20 cm/s. The larval remain close to the shore partly for this reason. Larvals with a length below 25 mm almost never swim more than 2 meters away from the shore. This can be explained by the fact that at 20-25 mm size the fish has
developed a more streamlined shape and can face higher velocities better. The larval Grayling does not have as specific preferences when it comes to water depth and substrate as the spawning Grayling. However, the closeness to shore and water velocities are of great importance (Nykänen, 2004).
The adult Grayling does not have as specific preferences as the spawning or larval Grayling. The observed preferences when spawning, appear to differ depending on location. This study is however focused on Grayling in Sweden, therefore data from Sweden or other Nordic countries have been prioritized and displayed in this section.
2.6.5 Restrictions and regulations
It is forbidden to catch Grayling in the entire Baltic Sea and the Bothnian Sea. In rivers and lakes the fishing of Grayling is limited. Only fish bigger than 30 cm in body length is allowed to be caught below the 67 º 00 north latitude and 35 cm above the 67 º 00 north latitude. The reason for this regulation is that it is desirable for the female Grayling to mate at least once before being caught. Fishing is not allowed between the 1st of April and the 31st of May due to spawning taking place (Eräluvat, n.d.).
2.7 Habitat changes caused by hydropower
When constructing a hydropower plant there are many impacts on the local environment. The obvious one is the physical properties that change by building a reservoir and thus adding a barrier to the natural water flow.
This does not apply to all hydropower plants, but many. The initial impact is great, but the vast majority of all hydropower plants in Sweden were introduced more than 50 years ago. Whether the local environment has reached a new "steady state" with the controlled flows of water or not, is up for discussion. It is argued that hydropower plants that were introduced many years ago have become almost like a natural part of the environment. The near-by environment have adapted to the new water flows and the constructed lakes (Fortum, 2015). These arguments are however often overruled with arguments from existing dam removal projects that have improved the life of vulnerable species, benefitted commercial fisheries, increased
recreational and cultural values (Headwater Economics, 2016). It is very hard to estimate exactly how large the environmental impact from a hydropower plant is. In this section different habitat changes that can be derived to hydropower will be explained. The habitats investigated are in consideration of fish.
2.7.1 Water temperature
The temperature of the water in rivers varies over the year with the seasons and the water cycle. As snow melts the flow increases and the water is cold. Over the summer the water in the reservoirs is heated up by the sun and as winter approaches it is cooled down. Hydropower and dams interfere with the natural
temperatures of the water, how much depends on the scale of the dam. There is a phenomenon called thermocline which is when layers in large bodies of water appear. These layers are caused by for example
19
different temperatures of the water since the density of the water differ depending on temperature. This is also linked to dissolved oxygen levels, which will be explained in the next section. When the water is layered up in a reservoir the water downstream from that reservoir is affected by what water that is being let out from the reservoir. If it is from the bottom the temperature of the water will be cooler than normal during summer and if it is from the top layer the temperature of the water will be warmer than normal. It thus depends on the construction of the dam how it affects the temperature (Fondriest environmental, 2014).
2.7.2 Dissolved oxygen levels
The level of free, non-compound oxygen is what is referred to when talking about dissolved oxygen levels. In other words the dissolved oxygen is the free oxygen molecules within the water. It is an important aspect and parameter when investigating water quality because aquatic life depend on it. Fish along with other species such as invertebrates, bacteria and vegetation all use the dissolved oxygen for respiration. The plants however only need it when there is no light to conduct photosynthesis. The dissolved oxygen enter the water via the air or as a by-product of plants. The oxygen enters the water most commonly by slowly diffusing across the water surface. It can also enter the water when air and water is mixed up under turbulent circumstances caused by wind, high waves, waterfalls or man-made systems such as hydropower plants. Different species have different requirements regarding dissolved oxygen. In Table 2 one can see a list of aquatic species and their minimum dissolved oxygen requirements (Fondriest Environmental, 2013).
Table 2: Oxygen levels required by certain species (Fondriest Environmental, 2013).
Oxygen requirements (mg/l) Species 7 Trout 6 Salmon 5 Bass 5 Carp 2 Grayling 2 Pike 1 Bacteria
Hydropower plants affect the dissolved oxygen levels in the river because when they release water through their turbines that water is taken from deeper levels of the reservoir and that water has a very low level of dissolved oxygen. So, typically water released from hydropower plants has a lower level of free oxygen compared to natural flows (Fondriest Environmental, 2016). The oxygen levels do vary depending on the hydropower plant and the size of the dam. Not all hydropower plants contribute to lower levels of dissolved oxygen. This phenomenon can partly be explained by low circulation of the water in still water such as lakes or large reservoirs. It can also be explained by stratification. Stratification is when the cold water with low oxygen drops to the bottom due to its density being higher than warm water with high oxygen levels. Since the water is taken from deep levels of the reservoir and then let out underneath the surface, without air turbulence, the low levels of dissolved oxygen spread throughout the river. However, currents, seasons and other aspects also have an effect on the dissolved oxygen levels. It is a complex system and it is hard to tell exactly which factors or parameters that contribute the most to low levels. It has however been observed that rivers with installed hydropower or dams have lower levels of dissolved oxygen compared to natural or un-exploited rivers (Fondriest Environmental, 2013).
20 2.7.3 Hydropeaking
The phenomenon of Hydropeaking, also known as short-term regulation, has been created by man. It is when the flow through a hydropower station is increased or decreased drastically because of fluctuations in the demand. Hydropeaking can have different time horizons. It can be on a yearly, weekly or daily basis. Over the year the demand is typically higher during the cold winter and lower during the warm summer. The week has fluctuations with higher demand during weekdays daytime when industries and other energy intense activities is going on and lower on demand on weekends. On a daily basis the production is low during the night,
increase fast in the morning, is fairly high during the day and decrease slowly during the evening and night. All of these production peaks has impact on the near-by environment, mainly because the increased flow flush away plants, fish and other organisms living in the river (Meile et al, 2008).
2.7.4 Stranding
One of the physical consequences of short term and possibly frequency regulation of hydropower, is stranding.
In opposite to Hydropeaking where the flow drastically increase and the fish can get flushed away, stranding happens when the water levels quickly decrease and the fish can become stuck on the shore. The decrease of water levels can be rapid and the fish risk not making it back to deeper levels and thus become stranded. The rapid decrease in flow and the following rapid decrease in water levels, can also be connected to the
Hydropeaking phenomenon. If the demand for electricity would drop fast this would be reflected by decreased water flows through the hydropower turbines (Saltveit et al, 2001). If the water level change rate is kept below 13 cm/ hour then the risk of stranding is reduced, says an often referred to study conducted in Norway on Salmon. Trout and Grayling are salmonid fish and this number can to some extent be assumed to be valid for Trout and Grayling as well (CEDREN, 2013).
2.7.5 Properties affecting stranding
A study done on juvenile Atlantic Salmon and Brown Trout in Norway shows that there are many factors that affect how likely fish may strand. One of these factors was the temperature of the water. Both juvenile Salmon and Brown Trout showed to be more frequently stranding during the winter when the water temperature was lower, than during summer when the temperature was higher. In this study the low temperatures were defined as 4,5 degrees Celsius and colder and the high temperatures were defined as 9 degrees Celsius or warmer. Apart from the seasons and water temperature, the light conditions also matter. During the daytime the fish, especially the juvenile Salmon, were more likely to be stranded compared to night time. This can partly be explained by the fact that both Salmon and Brown Trout are more active during the night. Another factor that affect stranding is if the fish has been cultivated/farmed or is wild. The Norwegian study showed that cultivated Salmon were more likely to become stranded than wild fish, the difference was particularly large during night time. Another interesting fact observed during this experiment is that many fish were reused during these tests, and even though they had been in the exact same situation before the fish did not learn how to avoid stranding (Saltveit et al, 2001).
2.8 Environmental quality standards in Sweden
The Swedish Agency for Marine and Water Management is responsible for the guidance of environmental quality standards (EQS) that concerns water quality and the marine environment. The EQS are legal
instruments that were established in 1999 as part of the Environmental Code. They were established with the purpose of coming to terms with emissions and other environmental impacts from diffuse sources such as traffic and agriculture. The standards should be designed in a way so that they describe the lowest acceptable quality or the desirable condition of the environment. The standards should be based on scientific criteria and
21
normally they apply to a geographical area. In Sweden municipalities or other governed agencies are responsible for compliance with the standards (SAMWM, 2016c).
2.9 Ecological status and potential
In order to be able to classify how well the Swedish environment is doing, ecological status and potential has been introduced. These are classification methods used to map the quality of the environment in larger areas.
The ecological status is applied to water flows that haven’t been altered or modified by humans. The ecological potential is applied to water bodies that have been heavily altered or modified. In rivers with hydropower plants it is therefore the latter, ecological potential, that is applied. The outcome of an ecological potential assessment can be one out of these five judgements: high, good, moderate, not sufficient or bad. The outcome is determined by quality factors. These factors can be divided into three groups: biological factors,
physiochemical factors and hydro morphological factors. These are then weighed together and represent the status or potential of the evaluated water body. It should be mentioned though that out of these factors, biological factors are the most important ones and they weigh heavier than the other two. The national goal is for all waters to obtain high or good status or potential (VISS-hjälp, 2010). The assessment if the water is heavily modified or not, determinate if the environmental quality should be status or potential for a water body.
2.9.1 Hydrological parameters
When evaluating the hydro and morphological factors there are certain parameters that are closely examined.
The hydrological parameters differ depending on water regime. Lakes, coastal waters and rivers have different parameters that are assessed. For rivers, which is the relevant water regime of this study, the following parameters are investigated:
Specific flow energy.
Volume deviation.
Flow change rate.
Water level change rate (VISS-hjälp, 2010).
Out of these parameters, the water level change rate is the one being investigated in this study. In Figure 3 below an overview of the system can be seen. There are many sub-categories in the HVFMS (Havs- och vattenmyndighetens föreskrifter) documents, however the path to understand the parameters explained in this section is highlighted in green in Figure 3.
22
Figure 3: Overview of the classification system. The green path leads up to the relevant parameter for this study.
2.9.2 Water level change rate
Stranding can be linked to the last of these parameters: water level change rate. The water level change for a river is calculated by the following equation:
𝑊𝑎𝑡𝑒𝑟 𝑙𝑒𝑣𝑒𝑙 𝑐ℎ𝑎𝑛𝑔𝑒 𝑟𝑎𝑡𝑒 [%] = 𝑎𝑣𝑒𝑟𝑎𝑔𝑒 (𝐴𝑏𝑠|𝐻𝑅𝑖− 𝐻𝑁𝑖| 𝐻𝑁̅̅̅̅̅ )
Where HRi is the regulated water level during the actual day, HNi is the unregulated water level and 𝐻𝑁̅̅̅̅̅ is the average regulated water level during the entire time period investigated. Abs stands for absolute value. The unit for the inputs in this equation is meters per hour. The outcome of the equation is a unitless value that is viewed as a percentage. The outcome is then assessed and classified according to Table 3 (HVMFS, 2013).
Table 3: How the status and potential are assigned (HVMFS, 2013).
Status / Potential Class Decrease or increase relative to the natural change of water level rate High 5 Deviation with maximum 0,05 m/ h
Good 4 Deviation between 0,05 and 0,15 m/h Moderate 3 Deviation between 0,15 and 0,3 m/h Unsatisfactory 2 Deviation between 0,3 and 1 m/h
23
Bad 1 Deviation with more than 1 m/h
24
3.0 Aims and objectives
This section will present the research question that this study tries to answer, the main aim of the study and the three objectives for the study.
3.1 Specific research question
How does the primary regulation at a given* hydropower plant affect the environment downstream with consideration of Trout and Grayling (Salmo trutta and Thymallus thymallus) and stranding?
* The location will be known within Fortum but secret to people outside the company.
3.2 Aim
The overall aim of this thesis is to understand if fish is affected by frequency control at a given location and if yes, how they are affected.
3.3 Objectives
The objectives of this study are:
Evaluate the stranding risk for Trout and Grayling and the effects of normal frequency regulation in the studied river reach.
Evaluate the stranding risk for Trout and Grayling and the effects of extreme regulation in the studied river reach.
Discuss the frequency regulation and its effects from both a global and local environmental perspective.
3.4 Project plan
This study has taken place during the spring semester of 2017. Below, in Table 4 a simplified time plan can be seen that describes deadlines throughout the semester.
Table 4: The project plan for this study.
Date Deadline
16th of January Start of the project 3rd of February Project plan ready
17th of March Preliminary background written 31st of March Preliminary method written
21st of April Results written
10th of May Preliminary analysis and discussion written 10th of May Hand in draft of thesis report
29th of May – 16th of June Hold presentations of the report (Fortum & KTH) 29th of May – 16th of June Hold opposition for another student
After presentations Hand in final reports for review and approval
25
4.0 Method
This section will describe how the work leading up to this thesis was executed.
4.1 Literature review
This master thesis is partly based on a literature review. The used literature have consisted of both public documents but also material that is only for internal use within the companies of Fortum and Sweco. A complete list of references used in this report can be found under "8.0 References" and a discussion of the reliability and relevance of the references and data can be found under "6.7 Credibility of references" later in the report. The literature review is the basis of the theoretical background. It is also the basis for how the outcome of the model is linked to the habitat preferences of Trout and Grayling.
4.2 Interviews and correspondence
During the work with this report several people have been contacted via e-mail or telephone in order to contribute with their expertise on the subject. If a person has been used as a reference he or she can be found under “8.0 References” at the end of this report. Only people with good insight and knowledge about the specific subject have been contacted and the given information has been doubled-checked with online sources.
4.2.1 Field trip
During the early spring a field trip to the selected power plant and the downstream river area took place.
During this visit the areas that were assessed to be valuable fish habitats or spawning places were documented with camera. During this visit the conditions were the following: no active frequency control, 50 m3/s water flow and the visit took place before the spring flood.
4.3 Modelling
A model that describes and simulates the water flow is a large part of this study. It is used as one of the main basis for the results together with literature and interviews. In this section, the software, input data,
assumptions and validation of the model will be described.
4.3.1 Software
Two main softwares have been used in this project. The first program is ArcGIS with the added toolbar Hec- GeoRas. The second program is Hec-Ras. These programs and their properties will be described in this section.
A complete description of the workflow can be found in Appendix 1. There it is possible to follow the steps of constructing the model step by step.
4.3.2 ArcGIS
ArcGIS, short for Geographical Information System is a software that allows the user to interact and work with maps, spatial analysis and data. The program was created by Environmental System Research Institute (ESRI) based in California as early as in 1999. The version used in this project is 10.3.1 (ESRI, 2017). All transformations and selection of data before the simulation in Hec-Ras started, has been done in ArcGIS. The toolbar HEC- geoRAS was added to ArcGIS in order to label and export the data in a way that Hec-Ras should later understand.
4.3.3 Hec-Ras
The program that has been used to model this project is Hec-Ras and the version 5.0.3. Hec-Ras is short for Hydrologic Engineering Center’s River Analysis System and has been developed by The U.S. Army Corps of Engineers (USACE). The software has many functions but the main reason for it being used in this project is
26
because it can simulate one-dimensional steady flow calculations and one- and two-dimensional unsteady flow calculations (USACE, 2016).
4.3.4 Data input
The terrain data used as a basis in the model created for this study has been gathered by others as part of a project investigating the consequences of a dam break. This data has been gathered by laser scanning at an altitude of 3 000 meters and with a density of data points of one point per 0,2-0,3 m2. The water flow at the time of the laser scanning was approximately 70 m3/s. The height system used is RH70. The data was provided by the consulting firm Norconsult upon request by Fortum. The acquired dataset consisted of two raster files spanning over a large area. The areas of interest for this study were located in the overlap between these two raster files. The resolution of the data was that one pixel represented an area of 5x5 meters.
Due to the collection method of the data (laser scanning) the elevation data underneath the water surface is unknown. The water surface reflects the laser and thus the data for the river bed is just a flat area, much like the water surface. This data can be collected by other methods but there was no room for that within the frames of this study.
The data used to validate and calibrate the model can be reviewed under section “4.3.4 Calibration and validation”.
4.3.3 Assumptions
In order for Hec Ras to work, certain assumptions must be made about the water system. One of these assumptions is that the river has a constant geometry and that no morphological changes take place. This means that no erosion takes place along the river bank or on the bottom of the river. This assumption is made because the continuous change in the water system is too complex to take into account. Another relevant assumption is that the water is clean. This is also related to the no erosion assumption. However, this is assumed because water that contains other elements such as sand or similar might behave differently. The selected study area can be considered to contain clean water.
Certain aspects of the model has been assumed with help of experts in hydrology. These aspects are Manning’s N value that is used to describe the hydrological roughness of different terrains or materials. Deciding on the Manning’s N value was done in close cooperation with an experienced hydrological engineer at Fortum. A table with the selected values can be viewed under the result section.
The shape of the river bed has also been altered and in some extent assumed. Since this data was lacking due to the data collection method, it has been altered manually. The field trip to the studied location has been used as basis for the shape of the river bed. But even though the water was calm on the day of the field trip and visibility was good, it has to be viewed as assumptions. The river is very wide and the water was not clear, these factors makes it impossible to investigate the entire river bed from land. In Picture 1 the substrate near the shoreline can be seen. In Picture 2 the river size and shoreline can be seen.
27
Picture 1: A picture of the shoreline and river substrate from the study area.
