Inundation Modelling An example of Scientific Support to DG-ECHO - - PowerPoint PPT Presentation

Mosul Dam Break Inundation Modelling

An example of Scientific Support to DG-ECHO and UN

A Annunziato, I. Andredakis, P. Probst

Joint Research Centre European Commission

2nd Scientific Seminar of the Disaster Risk Management Knowledge Centre Rome 9-10 March 2017 - Italian Civil Protection

Table of content

• Background
• Analysis objectives
• Initial and boundary Conditions
• Results of the dam break cases
• Early Warning Systems
• Future activities
• Response to our analyses
• Conclusions

44 km

Mosul Dam, Irak

11 km3 of water 2 km dam width 110 m dam height

Historical persepctives

• The dam was constructed between 1981 and 1986
• It dams the river Tigris, forming an artificial lake that holds

11.1 km3 of water

• It is located 40km NW of the city of Mosul and its primary

function is to supply electricity to the city’s 1.7 million inhabitants and to support irrigation needs.

• It is the largest dam in Iraq, measuring 2 km in length and 113

m in height. It is an earth-fill embankment-type dam with a clay core.

The problems of the dam

• Since the dam is built on a foundation of gypsum,

anhydrite and limestone, water can seep under the dam and compromise its stability, posing the risk

• f catastrophic failure.
• To avoid this, continuous maintenance has taken

place since the dam’s opening, in the form of grouting: new leaks are plugged by injecting a cement mixture into the compromised spots.

• The grouting stopped in 2014 with the arrival of
• ISIS. Even if the dam was back in the control of

Peshmerga (North Kurdistan) late 2014, the machines for grouting were severely damaged.

• In Early March 2016 the Iraq Government

announced the signature of a contract with the Italian company Trevi that should restart the grouting operations, which are now ongoing.

• Nevertheless the risk of failure of the dam is still

present and the international Humanitarian Community decided to send an exploratory mission to Iraq

Analysis objectives (1)

• In April 2016, upon request of the OCHA Country Office in Iraq, an UNDAC

team and an Environmental Emergencies expert was deployed to support preparedness for a potential failure of the Mosul Dam.

• Through networks and partnerships supported by OCHA’s Emergency

Services Branch, the following support activities have also been activated:

• The World Health Organization deployed its Emergency Medical Teams’

manager to Iraq to develop a protocol jointly with the UNDAC team

• The European Union’s Joint Research Centre has, through the Union’s Civil

Protection Mechanism, developed a number of dam failure scenarios (including flood maps)

• UNOSAT has provided satellite imagery
• The Joint UNEP/OCHA Environment Unit has provided an overview of

hazardous industrial installations in the flood zones

• The main objective was to understand
• The determination of the inundation extent and timeline in the first 12 days
• The role of the initial lake height on the inundation extent and dynamics
• The influence of the controlled discharge into the Tharthar Lake
• Initial assumptions revised on the basis of requests coming from the mission
• n the field
• To estimate new conditions (i.e. level reduction to a specific height)
• To reply to objection by local institutions (i.e. influence of Tharthar Lake)

Initial and Boundary Conditions (1)

• Tigris River topography
• 90 m resolution SRTM data by the CGIAR Consortium for Spatial

Information, resampled to 180 m – http://srtm.csi.cgiar.org

• Lake modelling
• Use of the results of the 2011 survey campaign published in open

literature and digitized

• Dam Break
• Coarse modelling of the break, constituted by 15 nodes that are

considered open in the wall of the dam. 3 types of breaks considered: 7%, 26%, 60%.

Lake bathymetry (Issa, 2015)

67 30 20 40 60 50 10 Elevation a.s.l. (m)

Lake of Mosul dam Mosul dam Main section of the Mosul lake

Tigris river Mosul lake main section Mosul lake shore Mosul dam and plant area 321 254 Elevation a.s.l. (m) 340 240 330 300 290 280 270 260 250 310 320

Dam

Simulation of the break

The break is opening instantaneously, creating a critical flow at the edge of the water. The instantaneous opening can be considered a conservative assumption or a sudden explosive break

Initial and Boundary Conditions (2)

Initial and Boundary Conditions (2)

Initial and Boundary Conditions (2)

Lake Height: 330 m, break size 26%

Mosul Baghdad

Mosul: max. 26m

330 m lake height

Sensitivity Analysis on Lake height

0.00 50.00 100.00 150.00 295 300 305 310 315 320 325 330 335

Time of arrival (h) Lake Elevation (m a.s.l)

Arrival time vs. Lake Elevation

Mosul Banyji Tikrit Samarra Bagdad Linear (Mosul) Linear (Banyji) Linear (Tikrit) Linear (Samarra)

• Poly. (Bagdad)

Simulated lake heights:

330m, 319m (current), 315m (proposed), 309, 307, 305, 300m

Sensitivity Analysis on Lake height

10 20 30 295 300 305 310 315 320 325 330 335

Maximum height (m) Lake Elevation (m a.s.l)

Maximum height vs Lake Elevation

Mosul Banyji Tikrit Samarra Bagdad Linear (Mosul)

• Poly. (Banyji)

Linear (Samarra) Linear (Bagdad)

Simulated lake heights:

330m, 319m (current), 315m (proposed), 309, 307, 305, 300m

Current Lake Level (319m)

Population Impact: 4.5 million people

1.5 million in water > 2m 140 000 in water > 10m

Population affected

The effect of the discharge into the Tarhar Lake

• After the performance of the previous calculations, the

preliminary assessment by the UN team onsite with the local authorities pointed out that in the calculations the presence of the connecting channel between Samarra and the Tharthar Lake was not considered.

• The presence of this channel could, in principle, relieve the

amount of water coming from the Mosul dam reservoir in the direction of Baghdad, diverting it towards the lake and thus reducing the inundation of the Baghdad plane.

• According to local engineers the maximum flow in this channel is

in the order of 30000 m3/s, which however seems not realistic (i.e. Rio de la Plata 22000 m3/s).

It is not clear how the control of the water flowing to Tharthar lake is performed. Apparently no control structure appears on the channel to Tharthar lake, so the only control is performed with the dam towards Bagdhad

Channel modelling

Lake height: 315 m with diverting channel

• Minor influence of the

diverting channel

• The inundation of

Baghdad is not prevented by the channel

• The volumetric flow to

the channel is less than 3000 m3/s while the incoming flow is one

• rder of magnitude

larger

Critical infrastructures

Colour-coded Affected Chemical Sites

Early Warning Systems

• Objectives:
• Identify critical conditions for the dam
• Raise awareness on possible need for evacuation
• Alert that evacuation is necessary: avoid false alerts !
• Last Mile Communication: the information needs to reach the

people !

• JRC experience on Tsunami Alerting Systems &

Last Mile

• Monitoring Sea Levels
• Information & alerting panel

Alerting logic based on sequence of measured level alerts

The alerting could be given with 1-1.5 h lead time, which is rather large time if a proper preparedness of local people is performed Time Water Level

Control Box

Nice review of the study by Prof. Al Ansari

• … As a conclusion this study may be

considered as comparable to the SC in its comprehensiveness and end results. This study, however, surpasses all the previous studies even the SC study by estimating the possible number of affected population and giving indications of the amount of material destruction at each level

• f inundation.
• (SC=Swiss Consultant)
• …
• 1. The JRC dam break study being the

most up to date and most comprehensive study done so far, so it must be checked for any areas of improvements weather in the assumptions

• r the updating of population statistics. The
• utputs of this study shall also define

“Rescue Line” in all the threatened zones for each reservoir water level scenario. Contour maps indicating these rescue lines should be prepared and projected on the ground with participation from the government to mark these rescue lines

http://ltu.diva-portal.org/smash/get/diva2:982079/FULLTEXT01.pdf

Several references to our study from press

2016

Conclusions

• The flood wave will reach the city of Mosul in less than 2 hours,

with heights of more than 20m in around 3 hours and its maximum height at the city, 26m, after 6 hours.

• An area housing about 180 000 people, almost 10% of the

city’s population, would be exposed to a flood wave of more than 10 m

• The water will continue downstream, having still a very

significant height that is only slowly reduced and affect a total of about 6 million people. A total of more than 500 000 people (including Mosul) would be exposed to flood waves of more than 5 m.

• According to these simulations, in the case of Mosul, an

evacuation within 4 to 5 km from the river would be advisable.

Conclusions (cont)

• Our results are in good agreement with results of previous
• studies. Here we add a detailed time profile of the floodwave’s

height and an up-to-date estimate of the affected population using GIS. We also provide detailed maps of the potentially affected cities.

• The analysis of the presence of the connecting channel between

Samarra and Thartar lake indicates that the maximum relief flow through this channel, does not have a relevant influence

• The results show that the lake height reduction plays a very

important role in the reduction of both quantities and also in the times elongation so that more time is available for the population to evacuate from the possibly affected areas

• Early Warning Systems should be implemented
• Similar type of analyses can be conducted for other dams to

have a verification of the preparedness of downstream populations