F IRST WORKSHOP OF THE THEMATIC NETWORK ON SHP

EFFECTS OF ADDITIONAL P ARAMETERS ON DEPLETED STRETCH 1 SCOPE OF THE DOCUMENT In the following paragraphs we try to prove the thesis that the reserved flow downstream of a weir or a dam can be reduced - achieving anyway a better environmental result - if the effect of the following additional parameters is duly taken into account in the reserved flow calculation: • Length of diversion section • Tributaries • Slope • Riverbed structure The thesis will be proven through the description of a case study. The scope of the document is not to give a solution, but to stimulate discussion among the major key actors of the field and to provoke counterexamples aiming to the confutation of the thesis. 2 DESCRIPTION OF THE PLANT The plant is located in a Alpine area of Northern Italy. The hydrological regime can be described by the following Flow Duration Curve not directly obtained by flow rate gauging on site, but by synthetic curves based on statistical handling of data obtained in similar near catchment areas. The characteristics of the plant relevant to our discussion are: Hydrological parameters: • Catchment area: 15,7 km2 • Mean natural flow 719 l/s • Reserved flow obligation 191 l/s1 • Percentage of reserved flow 26,5 % of mean flow • Residual catchment area between intake and tailrace 5,4 km2 • Average flow rate of tributaries 247 l/s Hydraulic parameters: 1

The reserved flow was imposed at 191 l/s because the plant is inside a natural park. The officers in charge gave no other reason or calculation to justify their assumption. The method applicable in the region so far gives the amount of reserved flow on the basis of the following formula: RF = 1,6 l/s/km2 ·1,8 (Precipitation factor) ·2,0 (naturalstic factor) ·S (km2, catchment area) = 90 l/s [12,5 % of the mean flow].

• • • • • • • • •

Length of diversion section Average riverbed slope Riverbed width (approx. rectangular section) Water depth with mean flow Wetted perimeter with mean flow Velocity with mean flow Water depth with reserved flow Wetted perimeter with reserved flow Velocity with reserved flow

Plant parameters: • Gross head • Plant rated discharge • Minimum plant discharge • Expected yearly production • Expected construction cost

2100 m 12.2 % 10 m 6 cm 10.11 m 1.3 m/s 3 cm 10.05 m 0.8 m/s

257.4 m 900 l/s 90 l/s 6,250,000 kWh 3,300,000 €

Alpine hydroelectric plant - Flow duration curve 6.500 6.000 5.500 5.000

Flow rate [l/s]

4.500 4.000

Mean flow Natural flow rates

3.500 3.000 2.500 2.000 1.500 1.000 500 0 0

30

60

90 120 150 180 210 240 270 300 330 360 Duration [days]

pag. 2/15

3

DO RESERVED FLOW FORMULAS WORK?

3.1 METHODS BASED ON VELOCITY AND DEPTH OF WATER The case study deals with a typical alpine torrent, with highly irregular cross section shape, a lots of big stones, cobbles, pools, natural weirs, artificial dykes, so that the adoption of the methods based on velocity and depth of water (see the document on different methods to calculate reserved flow) can lead to great errors, for the reasons already pointed out in the document mentioned: • Slope and natural water pattern don’t enter in the calculation • Diversion length and effect of tributaries or abstractions stay unconsidered • Without river re-structuring measures, in wide rivers these methods give very high values of reserved flow. • Reasonable use only for particular kind of diversion section • In mountain torrents give unrealistic values of threshold water depth • Suitable only for particular typologies of rivers, transferability doubtful. In the specific case, to obtain the water depth required by those methods the reserved flow should be higher than the mean flow and the reason is obvious: a constant slope and normal flow scheme doesn’t fit the real conditions of a torrent which is all but a regular shaped river. 3.2 METHODS BASED ON PHYSIOGRAPHIC PRINCIPLES One of these methods is in force in Italy to calculate the reserved flow (see note n. 1). On the basis of the rule in force the reserved flow should be approx 90 l/s. The effects on the hydraulic and plant parameters of such a value are: Hydrological parameters: • Reserved flow obligation 90 l/s • Percentage of reserved flow 12,5 % of mean flow Hydraulic parameters: • Water depth with reserved flow 2 cm • Wetted perimeter with reserved flow 10,03 m • Velocity with reserved flow 0,6 m/s Plant parameters: • Expected yearly production 7.100.000 kWh • Increase in yearly production 850.000 kWh As you can see a correct application of rule in force, without subjective interpretation of the officers could lead to a remarkable increase in Renewable Energy production. At he same time, if we take a look to the hydraulic parameters, even if the schematisation adopted is too rough, as we already pointed out, the change in them from 191 l/s to 90 l/s is not so dramatic, so that the lower value could be adopted. 3.3

METHODS BASED ON HYDROLOGIC OR STATISTIC VALUES

pag. 3/15

These methods present a large variety of values and percentage of particular reference flow rate. In the specific case no flow duration curve obtained by multi-annual measurements on the river is available, so that the applicability of the method per se is doubtful. Anyway, taking as a reference value, rather usual in many different countries, 10-15 % of the mean annual flow, a reserved flow of approx. 75-100 l/s should be imposed. More or less the same result obtained adopting physiographic principles is obtained, so that the considerations made about hydraulic parameters in case of reserved flow can be repeated. 3.4

METHODS BASED ON MULTI- OBJECTIVE PLANNING TAKING INTO CONSIDERATION ECOLOGICAL PARAMETERS

If, for example, the PQI (Pool Quality Index) method is adopted (see the document on different methods to calculate reserved flow), as the river is very rich in pools, the highest values proposed in the method can be adopted (4,3 l/s/km2 of catchment area), obtaining a value of approx. 70 l/s 3.5 COMMENTS AND CONCLUSIONS • All the methods more or less converge to a value around 90 l/s • It’s strange that the “ecological” methods give the lowest value. • “Hydraulic” methods give unrealistic results in torrents (see Errore. L'origine riferimento non è stata trovata.) 4 WHICH IS THE VALUE OF A L/S OF RESERVED FLOW In Italy energy produced from Renewable Energy Sources has the right to Green Certificates, so that an overall energy price of approx. 0,12 €/kWh can be obtained. In the specific case, on the basis of the flow duration curve available, each l/s produce about 9.900 kWh/year2 . In purely economical terms this means a reduced income of 1.188 €/(l/s reserved flow). In more general environmental terms, the release of each l/s of reserved flow means the that the same energy must be produced by conventional energy with the following more pollutants emission (in kg/years): min

max

SO2

17,8

134,6

CO2

7.634

9.879

NOX

13,9

31,7

Particulates

2,0

13,9

Methane

13,9

17,8

2

The expected yearly production without reserved flow is 8.100.000 kWh, so that Erf=(8.100.000 – 6.250.000)/191 ≈ 9.900 kWh/year/(l/s reserved flow) pag. 4/15

A lot of many other benefits are connected with the production from Renewable Energy Sources, but it’s out of scope now to recall them. 5 SEARCH FOR AN OPTIMISED SOLUTION An optimised solution of the reserved flow problem should aim to: a) Achieve the best results for local environment. In this case, not entering in a difficult discussion on the best values of a set of biotic parameters, this means to have hydraulic parameters (water velocity and depth, wetted perimeter) in the range suggested by the different methods in force. b) Achieve the best results for global environment. This means to maximise the energy production from a small hydroelectric plant. c) Keep the investment in a small hydroelectric plant profitable. 5.1 ALTERNATIVES To eliminate any subjective or arbitrary consideration, we’ll try to propose a set of possible alternatives to solve the problem. Then we’ll use a multi-criteria analysis method to search for the best solution, that is the best compromise solution. Obviously we have just in mind the preferred solution (this is normal in any objective or any tentatively scientific approach – our mind is not a tabula rasa), so that we limit the following description to a limited number of alternatives even to keep the global frame still interpretable. The alternatives considered are: 1) The design situation as it is (reserved flow = 191 l/s). In short it’s «ACTUAL» 2) Reserved flow = 90 l/s according to law in force. «90_SIMPLE» 3) Reserved flow = 90 l/s according to law in force taking into consideration the following possible improvements: § works of riverbed structure modification to increase hydraulic parameters § effects of tributaries In short it’s «90_MODIFIED » 4) As alternative 3 but with reserved flow fixed at 70 l/s In short it’s «70_MODIFIED » The first two alternatives need no further comments. The third and fourth will be briefly described. 5.1.1 Effect of tributaries As the diversion length is rather long, as typical in high head plants, and rich in well defined lateral valleys, as usual where the substratum of rock is superficial, the contribution of tributaries to the flow rates in the diverted stretch is important and can’t be left out. In the specific case, where the residual catchment area between intake and tailrace is almost 35 % of the catchment area itself, the mean flow rate from tributaries (varying from 0 at the intake to 247 l/s at the tailrace) contributes to increase of 30 % the fixed reserved flow of 191 l/s. If the reserved flow is reduced to 90 l/s the effect is even more glaring. So, why this fact shouldn’t be

pag. 5/15

taken into account in the reserved flow calculation, in order to reduce it? In fact, especially in this case where the riverbed width is almost constant along the diverted stretch, the beneficial effect of tributaries on hydraulic parameters (water depth, wetted perimeter, velocity) is sure. 5.1.2 Riverbed structure modification Very often reserved flow disappears in the riverbed because of the width of the river is great or the geo-morphological structure is such that water has preferential sub-superficial paths. It results in almost no water visible in the river with bad both visual and ecological effect. Itâ€™s true that in a torrent like the present case study one, the rock substratum is superficial and water losses in the underground are usually low (anyway it generally depends on rock fractures and on geology), but anyway the river wide and the risk is that low water disperses. To improve that situation a modification of the riverbed structure can be made. It can consist, for example, in realising a low water riverbed with slope and width such that water depth and velocity are higher enough to get a good visual and ecological result. Alternate to this narrower river stretch can be realised area with stilling water (pools) to favour fish spawning. The riverbed structuring should be eventually made according to indications of a specialist and, obviously, preserving the natural aspect of the river. Some problems could arise during hard flood events, because the new riverbed structure could be damaged. Anyway riverbed structure modification can help in reaching the scope of obtain with a lower value of reserved flow the same environmental beneficial effects of a higher one, with a double benefit for the environment: the production of a higher amount of renewable energy and a reserved flow hydraulically (and therefore biologically) correct.

pag. 6/15

pag. 7/15

5.2 WHICH CRITERIA TO CHOOSE? In any decision process a set of criteria must be chosen to evaluate the different alternatives. In the specific case we choose: 1) More energy produced with respect to the highest reserved flow (191 l/s) [kWh/year – environmental criterion] 2) Emission avoided connected to the more energy produced [tons of CO2 - environmental criterion] 3) Net Present Value (NPV) of the investment. This synthetic economical parameter, defined as the sum of discounted yearly cash-flow, takes into account: § The cost of the investment, depending on the cost of the riverbed structure modification; § The yearly income from energy selling, depending on the amount of reserved flow released downstream of the weir. The higher the NPV is, the better the investment is. The following assumptions have been made: a. The evaluation of energy value has been made on the basis of market price of 0,12 €/kWh for the first eight years of operation and of 0,55 €/kWh for the following years. b. Discount rate for NPV calculation is set to 7,5 % c. Discount time: 30 years (duration of the right of water diversion in Italy) d. Cost of the riverbed structure modification: 50.000 €, based on 15 working days of two excavators, design and supervision cost, etc. e. Operation and maintenance yearly costs: 80.000 € in case riverbed structure modification, 70.000 in the other cases. 4) Water depth in case of reserved flow [cm – environmental criterion]. It’s assumed, in this early preliminary stage, that the riverbed structure modification consists in the realisation of a low water riverbed with constant slope equal to the natural one and a width of 3 m instead of 10 m of the natural one. In reality, the modification could consist even in a reduction of the slope along the stretch, concentrating in small jumps (which can be easily overtaken by fish) the remaining difference in ground elevation. Furthermore in the alternatives 90_MODIFIED and 70_MODIFIED the water depth is calculated for a flow rate equal to reserved flow plus 50% of the mean flow rate coming from the tributaries (approx. 120 l/s). 5) Water velocity in case of reserved flow [m/s – environmental criterion] 6) Wetted perimeter in case of reserved flow [m – environmental criterion]

pag. 8/15

5.3 MULTI- CRITERIA ANALYSIS To evaluate which is the best alternative a multi-dimensional multi-criteria method has been used. We don’t enter now in the mathematical details of the method, which can be found in the “Guide to environmental approach to small hydroelectric plants” prepared within a Thermie Project of the 5th Framework Programme. Multi-criteria analysis or, more precisely, concordance analysis, evaluates the outcomes of a set of alternative projects in terms of multiple decision criteria applied simultaneously. For each alternative project it must be possible to obtain a measure of its effects in terms of each individual criterion considered. To avoid any kind of arbitrary evaluation, such measures should generally be quantitative, and could take the form either of technical-related criteria or economic-related criteria or environmental-related criteria. In this specific case we deal only with environmental and economic criteria. In any case, the measure of each effect must be determined prior to performing concordance analysis. The objective part of the input data is represented by an effects matrix that contains the effect of each alternative project in terms of each decision criterion. In the specific case we limit to six parameters only. Here you can see both the effect matrix and the vectors of weights.

Each element of the matrix quantifies the effect of each alternative in terms of each criterion, which has a certain associated importance. Decision criteria may refer to positive effects (benefits) or negative effects (costs). High values are obviously to be preferred for the former, whereas minimum values are desirable for the latter. Estimating the weighting factors is a key operation because their values strongly influence the final result. In the example we assigned the same weight to each criterion even if it is rather unreasonable, whatever the evaluator standpoint is. Anyway this is only the starting point of an evaluation subject to the a sensitivity analysis. pag. 9/15

Weights can never represent an objective value because they are influenced by the decisionmaker's viewpoint of the real system. The criticalness of the weights evaluation calls for sensitivity analysis of both the adopted preference rankings and the values of individual effects. This is necessary to gauge the stability of the results when the input data are varied I don’t want bore you with mathematical details. Anyway, it is possible to calculate a set of indexes which consider separately both positive and negative aspects of the projects. The most representative is a final score which provides integrated information on positive and negative aspects of each plan, and can be named Global Synthetic Index.

The final decision should favour those alternatives with the highest values of the index. For the problem at hand, the final scores are given in the last column of the table. The preference for «70_MODIFIED » is clear, while the last position is assigned to «ACTUAL », which is together with «90_SIMPLE» the two alternatives resulting with a negative global synthetic index. «90_MODIFIED » is ranked in the 2nd position. We must keep in mind that the obtained results are not intended to provide an automatic solution to the decision-making problem. The underlying presupposition is that this method is an instrument for assisting decision-makers faced with complex situations, especially when the available data is hazy. As such, it must be used to perform sensitivity analysis, to ascertain the stability of the choices made. The method implemented involves aggregating groups of homogeneous criteria and varying the relative importance of various groups with respect to another reference group. In our example, we can identify three groups of criteria: Economical-related (NPV); Global environment - related (Energy production and emissions avoided) Local environment – related (water depth, velocity and wetted perimeter). The results of the sensitivity analysis can be easily seen in a diagram: in the following picture the diagram showing the projects with the maximum global synthetic index when the weight of each group of criteria varies is presented. The graph can be briefly commented:

pag. 10/15

•

ACTUAL and 90_SIMPLE aren't the best alternatives for any value of the weight, so that it should be anyway rejected. • 90_MODIFIED is the best one only for high values of the local environment criteria and low values of the economical one. • 70_MODIFIED is the solution which occupies the widest area of the triangle and, in particular the centroid, so that it could be said that it's the more well balanced. In the following picture the worst alternatives, that is the alternatives which have the lowest value of the Synthetic Index when weights vary, is shown.

pag. 11/15

ACTUAL is clearly the worst alternative in a very wide range of weights assigned to the criteria. We try now to see what could happen if we modify the ACTUAL alternative improving it by a riverbed structure modification similar to 90_MODIFIED and 70_MODIFIED alternatives and keeping into account the effect of tributaries. The previous graphs modify as follows:

pag. 12/15

pag. 13/15

The results are rather evident: the situation isn’t radically changed: • 70_MODIFIED still remains the best alternative in the widest range of weights assigned to the group of parameters. • ACTUAL enters the group of best alternatives in a narrow range of weights, that is only if the weight of local environment parameters have a weight higher than 70 % and economical parameters less than 34 % • ACTUAL still remains the worst alternative in the widest range of weights. 6 CONCLUSIONS In the quick overview given in the previous chapters we tried to demonstrate in a real case study that values of reserved flow imposed to small hydroelectric plants owners should duly take into account some parameters which allow for reducing reserved flow released downstream of the weir, if some “accompanying measures” are taken. In particular the modification of the riverbed structure and the consideration of the beneficial effect of tributaries along the diverted stretch have a very good positive environmental impact not only locally, but also globally if the reduction of reserved flow with respect to conventional methods of calculation is granted.

pag. 14/15

The result of the proposed approach is that the higher the value of reserved flow is, not necessarily the best the solution is for the environment, intending the environment in the widest possible sense. Moreover, we want only to point out the necessity to face the reserved flow in global terms, because we think itâ€™s dangerous to try to save the leaf forgetting the tree or the forest. With this example we donâ€™t want to be conclusive. On the contrary, we want to promote the discussion about this matter, inviting everyone involved in the small hydroelectric plants field to add positive and negative examples and to supply counterexamples or to confute the thesis proposed.

pag. 15/15