Sand Storage Dams Performance, cost-efficiency, working principles and constraints

Copyright by Josep de Trincheria Copyright by Josep de Trincheria Copyright by Josep de Trincheria Copyright by Josep de Trincheria Sand Storage Dams Performance, cost-efficiency, working principles and constraints De Trincheria, Josep, Hamburg University of Technology, Germany josepm.trinxeria@gmail.com Nissen-Petersen, Erik ASAL Consultants Ltd, Kenya nissenpetersenerik@gmail.com International Symposium on RWH in Ethiopia (ISRH) 9 th June 2015

2 Outline 1. Rationale 2. Study area 3. Methodology 4. 2015 s Results 5. 2012 s Results 6. Causes and factors Source: http://www.rpsrelocation.com/_borders/checklist.jpg 7. Practical recommendations

3 1. Rationale To implement sand dams showing satisfactory performance and cost-efficiency levels is a complex procedure There are key biophysical and technical constraints which limit their real-life performance and cost-efficiency Currently, these biophysical factors and technical constraints are not adequately understood

2. Study Area 4

2. Study Area: All sand dams evaluated 5

6 3. Methodology 2012: 31 SDs evaluated - On-the-ground physical survey - Geometrical factor, throwback; bimodal rainfall season; evaporation and seepage losses; baseflow from riverbanks; specific yield - Sand storage capacity; Yearly water yield; Supply capacity; Cost-Efficiency - Semi-structured interviews with direct beneficiaries 2015: 48 SDs evaluated (incl. 2012 s) - Identification sediment texture and spillway damage - Semi-structured interviews with direct beneficiaries water supply capacity during the dry season

4. 2015 s results 7

4. 2015 s Results: water supply during the dry season 8

Households 1978 1979 1979 1979 1979 1979 1985 1998 2006 2007 2008 2009 2010 2010 2010 2010 2011 2011 2011 2011 2011 2012 2013 2014 Months 1978 1979 1979 1979 1979 1979 1985 1998 2006 2007 2008 2009 2010 2010 2010 2010 2011 2011 2011 2011 2011 2012 2013 2014 9 4. 2015 s Results: water supply during the dry season 10 9 8 7 6 5 4 3 2 1 0 Water supply capacity during the dry season (households) 10 9 8 7 6 5 4 3 2 1 0 Water supply capacity during the dry season (months) 6% were supplying water during the dry season

4. 2015 s Results: Sand sediments 10

11 4. 2015 s Results: Sand sediments Sediments Coarse, 0% Soil, 6% Fine, 10% Silt, 42% Medium, 25% None, 17% 65% were not effectively accumulating sand sediments None were accumulating coarse sand sediments Medium and fine sand sediments were present in 35% of the SDs

4. 2015 s Results: Damaged sand dams 12

13 4. 2015 s Results: Damaged sand dams Robustness Wingwall 4% Leakage 6% Swept away 17% Spillway 2% None 71% 29% showed severe structural damages: washed away, leakage, wing walls, spillway

5. 2012 s results 14

2. Study Area: Evaluations in 2012 and 2015 15

Volume of sand (m3) 25 7 13 14 22 12 29 30 3 5 2 4 1 8 6 9 11 10 15 17 19 24 27 16 18 20 21 26 28 31 16 5.1 Sand storage capacity 8,000 Potential sand storage capacity 7,000 6,000 5,000 4,000 3,000 2,000 1,000 0 Sand storage dam identification 83% presented volumes of sand <1000 m3 7% of the SDs accumulated 58% Upper layers of highly mixed sandy, silty and clayey sediments Upper layers of silty and/or clayey sediments on top of a layer of sandy sediments

5.1 Sand storage capacity

5.1 Sand storage capacity 20

21 5.2 Specific yield of the reservoirs 35.00% 30.00% 35.00% 25.00% 20.00% 15.00% 10.00% 6.67% 5.38% 6.92% 7.14% 8.46% 20.00% 15.00% 10.00% 5.00% 3.00% 0.00% Average specific yield 6.9% 7 [3,12]% is the specific yield for silty and sandy clay alluvium sediments 5 times lower than 35% Cause: High proportions of silty and clayey particles in the reservoirs

22 5.3 Water stored in the sand Water in sand sediments No water (93%) Water (7%) 80% acted as conventional dams Surface water between 0,5-3 months after the rainy season

m3/year 23 5.4 Yearly water yield 2,000 1,800 1,600 1,400 1,200 1,000 800 600 400 200 0 25 7 13 14 22 12 29 30 3 5 2 4 1 8 6 9 11 10 15 17 19 24 27 16 18 20 21 26 28 31 Yield Yield seepage losses Yield evaporation Yield contribution riverbanks The average yields were 112 m 3 /year Total aggregated yield for the 30 SDs= - < 2 x higher minimum yield 1 SD - 6% of its minimum satisfactory capacity - 2 SDs responsible for 56% of the water yield Cause: low volumes of sand sediments + low specific yields

Losses (%) 26.05% 24 5.5 Evaporation 100.00% 90.00% 80.00% 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00% 25 13 7 14 12 29 3 5 2 4 1 8 6 27 24 15 19 26 28 31 16 20 21 18 % evaporation losses actual Evaporation and seepage losses were 50% High vulnerability to evaporation generalised shallowness of the reservoirs High depth of sediments is crucial Evaporation may uniformly affect at least the first 60 cm of coarse sand reservoirs

N households/dry season 25 5.6 Water supply capacity 90 80 70 60 50 40 30 20 10 0 25 7 13 14 22 12 29 30 3 5 2 4 1 8 6 9 11 10 15 17 19 24 27 16 18 20 21 26 28 31 Actual Evap H Actual Evap C Actual Riv H Actual Riv C Total aggregated supply capacity was 64 and 39 households This is 320 and 195 individuals 17,000 inhabitants in the entire study area 660 inhabitants is the typical village size The supply capacity may decrease during poor rainfall years and droughts when the study area may shift to only one rainy season per year Low resilience to poor rainfall years and droughts

EUR 26 5.7 Construction costs 22,500.00 20,000.00 22,313 17,500.00 15,000.00 12,500.00 10,000.00 7,500.00 5,000.00 2,500.00 0.00 949 Total costs EUR 241,899 Average costs EUR 8,639/SD [EUR 414,672 for 48 SDs] 5 times higher from 2006-2012 than 1978-2005 1978-2005 Low spillways, rubble stone masonry, low construction volumes 2006-2012 High spillways, reinforced rubble stone masonry, high construction volumes

EUR/m3 27 5.8 Cost-efficiency: Yearly water yield 22,500.00 20,000.00 17,500.00 Actual cost-efficiency evaporation Actual cost-efficiency contribution Costs 17,570 15,000.00 12,500.00 10,000.00 7,500.00 5,000.00 2,500.00 0.00 25 7 14 13 12 29 3 5 2 4 8 1 11 9 6 17 10 19 15 24 27 18 20 21 31 16 28 26 15 Average 5,402 EUR/m3 EUR 134,830 were invested in SDs producing yearly yields lower than 1 m3/year

EUR/household 28 5.9 Cost-efficiency: Water supply capacity 22,500 20,000 17,500 Actual average Hot Actual average Cold Costs 22,313 15,000 12,500 10,000 7,500 5,000 949 2,500 0 25 7 14 13 12 29 3 5 2 4 8 1 11 9 6 17 10 19 15 24 27 18 20 21 31 16 28 26 Average cost-efficiency was 6,768 EUR/household Between EUR 167,715 and 190,425 did not supply water to any household (70-78% of the total investment)

6. Causes and factors 29

6.1 Incorrect siting procedure Reservoirs with low volumes of coarse sand 30

31 6.2 Incorrect siting: Reservoirs with fine sand Specific yield of fine sand is 21 [28, 10]% well-sorted, lab conditions However: - Low sorting degree levels Ill sorted fine sand is 13 [25, 4]% - Layers of finer alluvium sediments (interbedding) - Partial saturation of the reservoirs during the first 5-10 years (Wipplinger, 1958) - Partial saturation during poor rainfall years and droughts Real-life specific yield of fine sand lower than 10% Higher probability of low performance and cost-efficiency levels in fine sandy reservoirs Not suitable sites

32 6.2 Incorrect siting: Reservoirs with fine sand http://www.geologycafe.com/images/sorting.jpg http://www.scientificsoftware-solutions.com/softwaregallery/pm8_img_39.jpg

6.3 Incorrect siting: Reservoirs with silty and/or clayey sediments 33

6.3 Incorrect siting: Reservoirs with silty and/or clayey sediments 34

6.3 Incorrect siting: Reservoirs with silty and/or clayey sediments 35

6.3 Incorrect siting: Reservoirs with silty and/or clayey sediments 36

37 6.3 Incorrect siting: Reservoirs with silty and/or clayey sediments

6.3 Incorrect siting: Reservoirs with silty and/or clayey sediments 38

6.4 Incorrect siting: Seepage losses 39

6.4 Incorrect siting: Seepage losses 40

6.4 Incorrect siting: Seepage losses 41

6.4 Incorrect siting: Seepage losses 42

6.4 Incorrect siting: Seepage losses 43

44 6.5 Evaporation losses Evaporation may have been undervalued and should be systematically taken into account 10% at 60 cm, 0% at 0,9-1,0 m, but for coarse sand sediments Higher depths with fine sand, and silty and clayey sediments! Shallow reservoirs are vulnerable evaporation up to 0,6 m along all the surface!

45 6.6 Incorrect structural design: One-stage high spillways Design principle: Maximise the accumulation of coarse sand Height of the spillway is a key parameter adequate flood velocities allowing deposition of sand sediments in as much different biophysical, rainfall, runoff and sediment transport conditions as possible Lower probability of deposition of fine grain-size sediments Avoid reservoirs formed by mixtures of coarse sand mixed with finer sediments low specific yield

46 6.6 Incorrect structural design: One-stage high spillways Solution: Build the spillway by stages of reduced height (-) Not adequate infiltration need to be taken into account (-) Construction time and higher costs for rivers with continuous water flow during the rainy season (intermittent) (+) Maximisation of the performance and cost-efficiency, robustness to variability of rainfall, flow and sediment transport

47 6.3 Incorrect siting: Reservoirs with silty and/or clayey sediments All sizes of silty, clayey, sandy and gravel alluvium sediments. Silt- and clay-like particles as suspended load at upper part of the water column. Heavier sand and gravel particles as bedload at the lower part of the flood water

6.3 Incorrect siting: Reservoirs with silty and/or clayey sediments 48

49 Source: Nilsson, 1988 6.6 Incorrect structural design: One-stage high spillways

50 6.6 Incorrect structural design: One-stage high spillways There are sand dams built in one stage with no apparent siltation problems Catchment areas producing large volumes of sand sediments and/or highly specific rainfall, runoff and sediment transport conditions However, it should not be assumed that these highly specific conditions are applicable to other areas Even in these cases, there is a high probability that the performance and cost-efficiency is not maximum - Variability of rainfall, sediment and runoff - Forced interbedding - Non-selective scouring - Maturity paradox

51 6.6 Incorrect structural design: Variability of the rainfall, runoff and sediment transport Annual rainfall at Dwa plantation in Kibwezi, 1927-1997 Seasonal rainfall variability at Dw plantation in Kibwezi, 1927-1997 Ten-year distribution of good, normal and poor rainfall years in Kibwezi

52 6.6 Incorrect structural design: Forced interbedding Construction of one-stage high spillways Higher probability to block the bedload and suspended load Production of normally graded-bedded reservoirs The expected water yield of the sand dam may be assumed to be much lower than the real-life water yield of the reservoir In Keren (Eritrea), one-stage high spillways have been systematically built in catchment areas producing large volumes of coarse sand The reservoirs showed a layer of at least 1 m of silty alluvium sediments below a layer of sandy alluvium sediment of 3 m depth

53 6.6 Incorrect structural design: Forced interbedding Systematic change in grain or size from the base of the bed to the top Normal grading with coarser sediments at the base, which grade upward into progressively finer ones

54 6.6. Incorrect structural design: Selective scouring of fine grain-size sediments Subsequent floods should not be assumed to systematically wash away fine sediments and leave the coarsest ones The selective scouring of fine grain-size alluvium sediments depends on the energy of the river flow The energy of the river flow is highly variable inherent extreme temporal and spatial variability of rainfall, flood and sediment transport The stage height reduces the energy of the river flow and makes it vulnerable to accumulate fine sediments in a wider range of conditions The depth of the fine grain-size sediments accumulated causes that the energy required to effectively scour is higher As higher the stage height, the lower the probability that the river flow will have the required energy to effectively scour silty and clayey sediments

55 6.7 Incorrect structural design: maturity paradox Towards the end of the filling up of the SDs built as one-stage high spillways Spillway height will always have the adequate height to only retain the coarsest bedload sediments The velocity of the flow has been increased as a result of shallower water depth (as it is intended by stages) Independently of the actual distribution of sediments in the reservoir, it will always appear to be homogenously filled up with sandy alluvium sediments Reservoir will be wrongly assumed to be mature The real-life specific yield may be significantly lower as expected low performance and cost-efficiency

7. Practical recommendations 56

7.1 Practical recommendations: Systematic siting procedure 57

58 7.2 Practical recommendations: Real-life specific yield Specific yield analysis of representative sediments in the reservoir should always be systematically carried out In the absence of those: It is recommended to use 20-25% as the reference maximum specific yield of coarse sand reservoirs

6.6 Incorrect structural design: One-stage high spillways Stage 1. Spillway is 50 cm above the sand level Stage 2. Flood has brought sand to the level of the spillway Stage 3. Spillway is raised to 50 cm above new sand level Stage 4. Flood has deposited sand to the new level of the spillway Stage 5. Spillway is raised to 50 cm above new sand level Stage 6. This procedure is repeated until the spillway is fully closed 59

60 7.4 Practical recommendations: Height of the stages Without an adequate evaluation of the bedload and suspended load characteristics, the precautionary height of 18-50 cm - 18 cm high proportions of fine alluvium sediments - 50 cm high proportions of coarse sandy alluvium sediments. The height of the stage should be adapted to the most probable minimum flow, i.e. the minimum height of the bedload Only after an evaluation of the rainfall, runoff and sediment transport, the implementation of higher stage heights may be possible

61 7.3 Practical recommendations: Multi-stage construction Even by stages, 100% perfection cannot be achieved with smallest floods the velocity of flow may always be low enough for the deposition of fine grain-size sediments of reduced specific yield. Spillways by stages according to the most probable flood SDs accumulating silty and clayey alluvium sediments during poor rainfall years and droughts. High spillways in one stage systematic accumulation of large volumes of fine grain-size sediments low performance and cost-efficiency

62 7.4 Practical recommendations: Naturally deep layers of sand sediments and/or alluvial aquifers

63 7.5 Practical recommendations: Increase storage capacity Clear relationship between the capacity of a dam and the spillway height. Direct impact on the construction costs and costefficiency, technical complexity and the robustness of the structure Need to take into account other relevant factors: - The slope of the riverbed and its effect on the throwback - The texture and depth of the sediments of the original riverbed free sand storage capacity and robustness to evaporation Higher specificity of suitable sites Lower replicability and transfer potential

64 7.7 Practical recommendations: Other alternatives Other rainwater harvesting storage technologies may show higher levels of performance and/or cost-efficiency Sub-surface dams, earth dams, ponds, etc.

Thank you for your attention! 65