Project Title: Design Options of the Flood Mitigation Plan of Sg. Muda, Sungai Muda, Kedah, Malaysia

Funder: Department of Irrigation and Drainage (DID) Malaysia

Duration: 1st April 2006 - 30th September 2006



Main Report (ISBN : 983-3067-17-4)



Prof. Dr. Aminuddin Ab. Ghani <>

Prof Dr. Nor Azazi Zakaria <>

Prof. Pierre Y. Julien (Colorado State University, USA) <>

Assoc. Prof. Dr. Rozi Abdullah <>

Chang Chun Kiat <>

Rosmaliza Ramli <>

Joseph Dinor <>

Asnol Adzhan Abd. Manap <>
Mohd Fazly Yusof


Project Summary:

The Flood Control Remediation Plan (FCRP) has been reviewed with the perspective to possibly enhance the proposed design.   The objectives of this review are: (1) ensure that design cross-sections and alignment of the main river channel are economic, effective and environmentally sound; (2) propose alternative designs for identified locations to meet the above requirements; and (3) examine the long term river behaviour through model studies to minimize expensive repair works in future resulting from the new alignment.

The scope of work of this report includes a review of:

(a)  Rainfall data in space and time leading to chosen flood events.

(b)  Runoff discharge and stage records of the chosen events;

(c)  Computer simulation of the flood events;

(d) Computer simulation of long term river behaviour to determine stretches prone to meandering, hence needing extra protection;

(e) Changes in alluvial river geometry in terms of aggradation and degradation as well as lateral channel migration as a result of the flood events;

(f)  Design criteria used for the Flood Control Remediation Plan;

(g) Specific design of the proposed structures including levee protection, riverbank protection works and protection of bridge crossings and other structures.

In an itemized fashion, the main conclusions of this review are specifically:

(a) There is significant uncertainty in the determination of the three-day rainfall rates in 2003.  The watershed average rainfall precipitation is difficult to ascertain from only four raingages.  However, the three-day watershed average rainfall precipitation is found to be less than 300mm. 

(b) The analysis of a 44 year measured discharge record at Ladang Victoria leads to a well defined 50-year flood peak discharge around 1,270 cumecs.  Similar results are obtained from independent flood frequency analyses.  This value is the most reliable estimate of the 50 year flood peak discharge.  The 2003 flood with a peak discharge of 1,340 cumecs is recommended for the design of the FCRP.  Comparatively, river stage records are highly variable as a result of excessive sand and gravel mining operations on Sg. Muda.  Stage records should not be used to estimate flood discharges. 

(c) Computer simulations of the flood events with the model HEC-HMS showed very good results for the calibration and validation of the discharge hydrographs at Syed Omar in 2003 and 1998 respectively.  However, there is a significant variability (error of at least 100 cumecs) in both the calibrated and validated results at Ladang Victoria.  This appears to be due to the limited number of raingages in the southern part of the watershed where precipitation showed spatial variability.  When using the same high rainfall precipitation as JPZ, the HEC-HMS results are very consistent with their design hydrographs.  However, as discussed above in (a), the three-day rainfall precipitation of 350 mm used to determine a 50 year peak discharge of 1,815 cumecs is viewed as excessively large.  The results of these rainfall-runoff models are highly variable and less reliable than the measured flood discharges at Ladang Victoria.  It is concluded that the 50-year design hydrograph with peak discharge of 1,815 cumecs is excessively large.  The value of peak discharge of 1,340 cumecs should be used for the design.  The design discharge reduction by 25% should results in significant cost savings for the FCRP of Sg Muda.

(d) The model Fluvial 12 has been used for the lateral and vertical elevation changes during the 2003 flood event.  Most cross sections are very stable and the vertical elevation changes are small due to the low sediment transport rates of the bed material.  The sand and gravel mining operations are far exceeding the natural sediment transport rates of the river.  These operations can cause serious problems to the stability of the proposed structures of the FCRP.  It is concluded that in-stream sand and gravel mining operations should not be allowed. 

(e) The model Fluvial 12 also highlighted potential lateral migration changes near the sharp bends upstream of Merdeka Bridge and near Lahar Tiang.  It is concluded that channel realignment near Merdeka Bridge and Lahar Tiang would minimize potential lateral migration problems of Sg Muda.  These realignments should also result in better flood conveyance, reduced roughness and backwater and lower bund elevation.

(f) The proposed bund height under Alternative 2 of JPZ appears to be highly conservative and will pass flows with period of return much in excess of 100 years.  The proposed design criterion for the FCRP is to add a one meter freeboard to the flood stage calculated for the current channel cross sections, without widening, using the 2003 flood peak of 1,340 cumecs.  The results obtained from the model HEC-RAS result in lowering the proposed bunds by 30cm to 1m over the 43 km reach of Sg. Muda. 

It is recommended not to widen the channel cross sections.  Channel widening will result in lower flood stages and lower bund elevations, thus resulting in additional cost savings.  These cross sections were not available to us, but further analysis should be carried out if channel widening is retained as an option.  In view of excessive sand and gravel mining of the Sg Muda, the bridge pier footings at Ladang Victoria need serious and immediate attention.  Pumping stations for irrigation canals have also been required on the Sg Muda as a result of the continuing degradation of the river.  The degradation can be clearly attributed to the excessive sand and gravel mining of the Sg Muda.  For the bank stabilization structures of the FCRP to be effective, it will be essential not to allow any in-stream sand and gravel mining activities in the future.  It is proposed to allow off-stream sand and gravel mining activities at a minimum distance of 50 m from the river banks.  The floodways to Sg. Merbok is not recommended because it would bring additional water to areas that were readily flooded in 1998 and 2003.  The Merbok river diversion would also require long bunds and an additional barrage, which cost will not be economically justified.  The potential diversion to Sg. Korok can only be justified as an emergency measure to prevent rupture of the proposed bunds designed in this FCRP.  These diversions may alleviate the problem of flood mitigation on the Muda, but at the same time may exacerbate the flooding problems in the upper watershed of Sg Perai.



The Flood Control Remediation Plan (FCRP) has been reviewed with the perspective to possibly enhance the proposed design.   The conclusions of this report demonstrate the potential for significant cost saving features as well as structural stability measures.  The specific recommendations for the FCRP include the following:

(a) It is proposed to reduce the design discharge from 1,815 cumecs to the proposed discharge of 1,340 cumecs from the 2003 flood.  This design discharge reduction of about 25% should yield significant cost reduction of the total cost of the FCRP.  It is recommended not to widen the channel.  However, any widening of the channel cross sections would result in further lowering the bund elevation thus further reducing the cost of the FCRP.  It is recommended to run the HEC-RAS model with the proposed widened cross sections for comparison with the current results (these cross section were not available for this study).  More specifically, the effects of the excessive channel widening downstream of the Muda barrage should be considered in terms of bund height and also possible sediment accumulation near the mouth of Sg. Muda.  

(b) The suggested channel realignment is expected to result in reduced roughness, reduced backwater effects and thus lower bund elevations.  It is recommended to run the HEC-RAS model with consideration of the proposed realignments.  The results of these two measures should result in a lower cost of the FCRP.

(c) It is strongly recommended to prohibit all in-stream sand and gravel mining activities between Ladang Victoria and the Muda Barrage.  Off-stream mining activities should instead be allowed at a minimum distance of 50 m from the channel.  We recommend a more detailed analysis of the locations and volumes of sand and gravel that can be extracted under this proposal. 

It is recommended to recalculate the bund height and provide a more detailed estimate of the cost saving features discussed in (a), (b), and (c).


Study Area: Sungai Muda Lower Reach                Sungai Muda Catchment

New Barrage Under Construction
Bed Load Sampling  @ Ladang Victoria New Bridge

(a) Calibration                               (b) Validation
Runoff Hydrograph at Ladang Victoria Discharge Station


Longitudinal Flood Profile for Sg Muda in HEC-RAS Model


Cross Sections for Sg Muda in HEC-RAS Model


Comparison of 2003 Flood Water Level and Proposed Bund Height

(a) CH 23.0                                                      (b) CH 21.9 
Lateral Migration @ Kg Lahar Tiang (FLUVIAL-12 Model)


(a) CH 15.4                                                      (b) CH 13.1 
Lateral Migration @ Bumbong Lima (FLUVIAL-12 Model)