It, has become fashionable to blame “cancellation” of Arun III project, by the World Bank in 1995, in reference to current electricity crisis, inferring that had it been implemented consumers in Nepal would not have been facing current load shedding problem. Now is time for a dispassionate and detached analysis of the two scenarios, the Arun III route and the route sans Arun III. The analysis of alternative scenarios demonstrates that the implementation of Arun III 201 MW project in 1995 would have not eliminated load shedding from 2000 onwards; in fact the power deficit would have been of a higher magnitude. Now, Arun III planning is for a new incarnation with a capacity of 402 MW, expected to be completed within next 10 years. But as the output of this project has been slated for export, it, too, will not alleviate load shedding problem of Nepal.
Key words: Arun III project, load shedding, Nepal
Nepal is facing severe electricity crisis due to supply constraint and it has already been forecast that the load shedding is here to stay in the increasing magnitude. During this year’s wet season Nepal faced the problem of flood which resulted in loss of life and limb as well as property in the hilly areas and Terai in Western Nepal. Besides, the breach of Koshi embankment played a havoc in East Nepal where people got drowned (and unfortunately killed, too) and displaced, homes washed away, and farmers’ investment of time, energy and resources in the cultivation of their land washed out (in some areas even fertile land was converted to sandy “beach”). Moreover, communication network and infrastructure including road network was brought to disarray. Industries in the area also suffered due to break down of transportation system; either because of failure to access raw materials and other supplies or because of inability to supply finished product to its customers (most of the industries suffered on both counts).
In the past, Nepal Electricity Authority (NEA) used to trot out the excuse of no water in rivers whenever it came up with a new load shedding schedule or made changes in them resulting in the increased hours of load shedding. However, this rainy season, even in the abundance of water (or flood!), NEA imposed a load shedding of 2 hours each day, two days a week, till third week of August 2008. With effect from 27th August, 2008 the load shedding hours was increased to 16.5 hours/week which kept on increasing ad nauseaum. The Chief of Load Dispatch Center of NEA announced in a program in Butwal on 11th November 2008 that there will be load shedding at the rate of 10 hours/day in Marga (November-December), which will go up to 12 hours/day in Paush (December-January) and it will peak at 14 hours/day during Magh (January-February) this year.
Many a pages have been written wherein the root cause of Nepal’s electricity crisis has been ascribed to the cancellation of Arun III by the World Bank in 1995, scheduled to be completed in 2005. Besides the journalists of many hues, a number of luminaries have also “invested” time and energy to write on this topic, including a special chapter dedicated to this subject, forming a part of the book by a renowned economist, who has occupied the position of finance minister of Nepal a number of times over last two decades and was the first Vice Chair of National Planning Commission of democratic Nepal in 1991. It is high time to examine/analyze if there is any truth in it. Irrespective of whether there is an iota of truth in this logic or not, it is also time to draw lessons from this phenomenon so that Nepal is able to learn from this episode and undertake projects for implementation in future prudently. In other words, a dispassionate and detached analysis is called for to test if the argument attributing current load shedding problem to the cancellation of Arun III project by the World Bank is rooted to ground reality or not and find out what lessons could be learnt from this episode. This entails conducting an autopsy of this logic, as it will also be able to throw ample light on the intellect and mindset of the hydrocracy (intelligentsia, politicos and bureaucracy involved in water resource sector) in Nepal.
A Decade’s Achievement since 1995
Let’s begin by taking stock of the situation with regard to the implementation of hydropower projects in a decade since 1995, subsequent to the cancellation of Arun III:
Table 1: Achievement through till 2005
Anticipated Achievement under Arun III
Had the World Bank insisted on continuing with Arun III, NEA would not have been allowed to build both Kali Gandaki A (144 MW) and Modi (14.8 MW) projects, nor would IPPs (including Chilime Hydropower Co. Ltd. – a subsidiary of NEA) have been allowed to build projects with more than 10 MW capacity due to the covenant imposed by the World Bank in the loan documentation of Arun III, restricting Nepal from implementing projects of capacity bigger than 10 MW. The covenant stipulated that “NEA … would seek IDA’s specific consent before undertaking any investment projects that would increase generation capacity by more than 10 MW capacity or transmission projects costing more than $ 3 million. In granting such a consent, IDA would need to be satisfied that the project is economically and technically justified and part of the LCGEP, that NEA has the financial and managerial capacity to undertake the project without delaying implementation of the Arun III or the rest of its ongoing program, that the project is consistent with HMG’s public expenditure program and macroeconomic framework…” [Financially too Kali Gandaki A would not have been built had Nepal gone Arun III route as fund earmarked by the ADB for Arun III was used for the former after cancellation of the latter.]
In this manner, if the country had started implementing Arun III at that time, the achievement, constricted by the covenant imposed by the World Bank, during that particular decade would have been as detailed in the following table:
It is now obvious that if Nepal had taken the Arun III route, only 220 MW capacity would have been added to the system in the decade ending in 2005, including Arun III, with an average annual generation of 1,845.86 GWh at the total cost of $ 1,130.77 million, with the average cost working out to $ 5,143 per kW, to be completed in 5.17 years in an average.
Forced by the World Bank to abandon Arun III, NEA added 165 Megawatt (MW) at the cost of $ 425.7 million – a decrease of 36 MW in the incremental capacity at a saving of $ 656 million. This indeed is a substantial saving for sacrificing 36 MW.
Although the average annual generation would have been higher by 52.5 GWh had Arun III been implemented, but by abandoning Arun III, Nepal not only succeeded in adding incremental capacity of 73.8 MW but also succeeded in saving $ 400 million in total (by the concerted efforts of NEA and the private sector). Compared to an average cost of $ 5,142.6 per kW, it actually cost only $ 2,485 per kW in an average for the capacity added during the decade. In terms of time too, Nepal gained as it took only 4.05 years in an average to add 293.68 MW compared to 5.17 years that would have taken to implement 219.88 MW.
Load Shedding Scenario
It is common knowledge that the load shedding is a function of the system capacity and electricity demand of respective years. For this purpose following tables compare the two scenarios comprising the system capacity and peak demand of the respective years, (a) based on actual commissioning of various projects and (b) anticipated commissioning of Arun III and other projects of up to 10 MW capacity in 2005. Table 3a below depicts the actual power surplus/deficit scenario during the period based on actual commissioning of various projects during that decade (year of commissioning of respective projects based on Table 1) and peak demand of those years:
From the perusal of the above table it can be seen that Nepal had an actual power deficit of 13.26 MW in 1999 and after commissioning of various projects, beginning with Khimti in 2000, the load shedding problem did get buried in the pages of history through till 2005. Actually the problem from 2000 till 2005 was one of “spill” (and the failure to encash such spill).
The Table 3b below analyzes the power surplus/deficit in the alternative scenario if Nepal had undertaken to implement previous incarnation of Arun III. It is a comparison of actual peak load in the respective years (same as in Table 3a) and actual commissioning of various projects less than 10 MW capacity as covenanted by the World Bank (same as in Table 3a except for exclusion of projects bigger than 10 MW) and anticipated commissioning of Arun III in 2005:
Table 3b: Peak Demand and System Capacity
based on implementation of Arun III scenario in MW
Above Table makes it abundantly clear that had Nepal chosen the Arun III route, the power deficit of 13.26 MW in 1999 would have snowballed up to 184.72 MW in 2004 and it would only have been reduced to 24.51 MW with the commissioning of Arun III in 2005 (assuming that it would have been commissioned in time).
By choosing a scenario sans Arun III, Nepal succeeded in avoiding an electricity crisis from 2000 through till 2005 and due to this route the magnitude of load shedding post 2005 is also relatively lower than had Nepal chosen to implement Arun III. In other words, until the completion of Arun III in 2005 Nepal would have continued to pass through severe and ever increasing load shedding problem, obtaining since late 90s, as only about 19 MW additional capacity would have been added to the system prior to 2005 (pending commissioning of Arun III) by a couple of projects of less than 10 MW capacity (e.g. 6.2 MW Puwa by NEA and some odd projects by IPPs totaling 12.683 MW).
Bitter Ground Reality
In order to get a correct picture of the alternative scenario, it also must be noted that there is no history of the public sector, donor funded, infrastructure projects in Nepal completing construction/erection as per the original schedule. Middle Marsyangdi project, slated to be completed in December 2004, is still under construction and is reportedly to be commissioned by December 2008. NEA officialdom defensively says that this particular project is a bad example to draw inference from. Therefore, one could attempt to draw interference from the much acclaimed Chilime project (which succeeded to garner a number of awards and laudatory comments) which did incur a time overrun of more than 5 years and cost overrun of Rs 800 million (included in the total cost of completion of Rs 2.33 billion) and it even had to change the horse (i.e. the civil contractor) in the middle of the race. It will be interesting to study the time overrun incurred by various projects since the days of Kulekhani I depicted in the table below:
With such a track record, Nepal would have been reeling under load shedding of a higher magnitude begun in late 90s, awaiting completion of Arun III which would have realistically succeeded to be commissioned around 2009 only. Specifically the power deficit of 24.51 MW in 2005 would have been 225.51 MW instead and higher by the same quantum in the following years till commissioning of previous incarnation of Arun III.
Potential Impact on NEA, Electricity Users and the Macro Economy
The impact of implementation of previous incarnation of Arun III would have been two fold; firstly due to unavailability of power during the period from 2000 through 2005 as per the analysis above and secondly due to higher cost per installed capacity.
The primary impact of implementation of Arun III would have been felt by NEA in terms of lost revenue due to power deficit (Table 3b) from 2000 through 2005. Cumulative power deficit from 2000 through 2005 would have been 552.74 MW (computation based on Table 3b) which is equivalent to 1,597.86 GWh at 33% plant factor (generally achieved by NEA in an average). The monetary value of such revenue loss, that NEA would have suffered, amounts to Rs 10.4 billion at Rs 6.53/kWh (average revenue rate of NEA for 2005).
Secondarily, and more importantly, the macro economy also would have suffered due to power shortage while Arun III was under implementation. A study on “Economic Impact of Poor Power Quality on Industry – Nepal,” conducted on the auspices of USAID-SARI/Energy Program has examined the cost of electricity supply interruptions both in terms of outages as well as loss of quality of power. Momentary interruptions, unplanned and planned outages, voltage fluctuations, and supply harmonics were considered. The conclusion relevant for the purpose of this article is with regard to planned outages. The industry-wide average cost of planned interruption is calculated to be $ 0.14 per kWh as depicted by the following table:
From this it is clear that due to the electricity deficit of 1,597.86 GWh during the period under consideration, the cost to the economy of Nepal would have amounted to a whopping Rs 15.66 billion due to the planned interruptions (at the rate of US 14 ¢ equivalent to Rs 9.80). The study referred to here has used the cost of operating standby electricity generation to arrive at the cost of planned generation as the main component of it. However, as even just the fuel cost of the stand by generator exceeds Rs 15/kWh, this estimated cost is basically cost to the industry which is on the lower side. If one was to compute the cost to the economy, the cost of the chain impact on the economy in terms of lost employment, loss of purchasing power resulting in loss of demand for goods, consequential loss of revenue by the Government of Nepal (GoN) etc. due to unavailable power will have to be taken into consideration. Therefore, the loss to the macro economy would have been exponentially higher.
The second set of adverse impact is related to the cost of implementation. With an estimated cost per kW of installed capacity of $ 5,143 it would have exerted upward pressure on NEA’s retail tariff. Moreover, the increase in retail tariff would also have been necessitated by the key covenants agreed by the Government of Nepal (GoN) with Asian Development Bank (ADB) under Seventh Power Project of ADB which requires NEA to ensure 6% return on assets (later revised to “revalued assets”), 23% self investment ratio and debt service coverage ratio of 1.2. Failing to revise the retail tariff upward would have forced NEA to incur massive loss. One needs to remember that, as NEA is fully owned by GoN, the loss suffered by NEA is tantamount to the loss suffered by GoN and, hence the adverse impact on the macro economy as well.
Besides, going the Arun III route with an estimated cost per kW of installed capacity of $ 5,143, the actual cost of implementation would have been a lot higher (perhaps it could have been, hopefully, capped at $ 8,000 per kW) as most of the donor funded projects have the trend of not only incurring time overrun, but cost overrun too, exerting further upward pressure on NEA’s retail tariff by a higher magnitude or, with no tariff increase the loss to be suffered by NEA would too have been at a higher plane and consequential adverse impact on Nepal’s economy would have been higher too.
New Incarnation of Arun III
GoN has executed a Memorandum of Understanding (MoU) with Satluj Jal Vidyut Nigam Ltd. (SJVNL) on March 12, 2008 for the implementation of Arun III hydropower project. The salient features of this project are as follows:
· Type : Pondage run – of – river
· Installed Capacity : 402 MW ( 6 x 67 MW )
· Design Discharge : 320 m3/sec
· Gross Head : 303.8 m
· Dam Height : 68 m
· Powerhouse Type : Underground
· Annual Energy : 2891 GWh
· Project Cost : $ 859 million at 1995 price level
According to the MoU, this project is being developed as an export-oriented project. Besides the capacity and energy royalty stipulated in the Hydropower Development Policy, 2001, SJVNL has also agreed to provide 21.9% power (88 MW) free of cost to GoN.
Interestingly, SJVNL estimates Arun III to cost $ 859 million (at 1995 price level) for 402 MW which works out to $ 2,137 per kW. It is educative to note that the aborted Arun III was going to cost $ 5,383 per kW as its estimated cost was $ 1,082 million for 201 MW. In other words, SJVNL cost estimate of $ 859 million for 402 MW is lower by $ 223 million compared to its previous incarnation with just 201 MW capacity (a tragicomic situation of lower capacity project costing more than higher capacity one). This impels one to conclude that Nepal was lucky to have 201 MW capacity Arun III project cancelled in 1995 as it has spared Nepal from another blunder in hydropower development. The new incarnation is going to be built at a reasonable cost of $ 2,137 per kW but, unfortunately for consumers in Nepal, the cheaper electricity is slated to be exported.
The apologists of previous incarnation of Arun III have been doing their best to justify the high per kW cost of $ 5,383 per kW by saying that the amount includes infrastructure development costs like construction of access road and construction/erection of transmission line. Since 1995 there has been no construction of access road and transmission line of any significance in the project area and, therefore, SJVNL is going to have to build the infrastructure and the cost estimate must include the cost of building such infrastructure.
Mitigation of Nepal’s Load Shedding on Commissioning of Arun III
From the content of the MoU, it is clear that this project is being developed by SJVNL as an export oriented project and, therefore, out of 402 MW only 88 MW will be available for utilization in Nepal and this project will be able to mitigate Nepal’s electricity crisis to that extent only by when Nepal’s demand would have risen significantly. According to the analysis conducted by the author of this article, Nepal will be facing severe load shedding even after 10 years by when GoN plans to add 10,000 MW to the system, including implementation of Arun III if GoN fails to comprehend the problem and get its act together SOON. Otherwise, even after commissioning of new incarnation of Arun III, Nepal will still be facing electricity crisis.
The electricity users in Nepal, suffering from the vagaries of the load shedding, are wondering why the new incarnation of Arun III is being developed as an export oriented project when Nepal is facing severe electricity crisis. There are certain projects on the anvil which are located in Far Western and Mid-Western Development regions of Nepal – too far from the load centers – and, therefore, evacuating electricity from there to the load centers (Central Development and Eastern Development regions of Nepal) is deemed unviable mainly due to high technical loss. Technically this is not a correct logic as the power from these areas could easily be evacuated using 400 kV transmission network. But the incremental cost of erecting high voltage transmission system might not be justified by the value of technical loss that will be reduced, mainly from the perspective of cost benefit analysis.
The above logic does not apply in the case of Arun III as it is located right at the hub of the load center, which is relatively highly industrialized and the industries in these regions are not operating at its full capacity due to lack of power. The industries in the area need additional 200 MW right now and by the time this project will be commissioned 400 MW will be consumed easily. Moreover, it is illogical for NEA to plan to mitigate future load shedding by importing power from India which will cost almost treble of the cost of electricity from Arun III. It would have been logical to use cheap electricity from this project in Nepal and export only the part that Nepal will not be able to use.
Lessons for Future
Ø Merely because the new incarnation of Arun III (402 MW) is estimated to cost $ 859 million, it cannot be said that the old incarnation (201 MW) should cost half of the new incarnation as the cost does not increase or decrease in a linear fashion. However, the cost estimate of $ 1,082 million for 201 MW is, obviously, inordinately high. It reflects very poorly on the competence and capability of those preparing the cost estimate of the previous incarnation. GoN and NEA should design a training program to build the capability of the technocrats involved in estimating cost as such and impart such training to them from time to time.
Ø The concerned authorities, in coordination with the academia, need to develop a norm as to the range of cost for hydropower implementation and for this purpose reference could be taken from other countries in the region that have experience in developing hydropower projects like India, China, Sri Lanka, etc.
Ø Luckily that Nepal didn’t get to implement a hydropower plant whose estimated cost per kW was more than $ 5,000 while average cost of hydropower projects implemented so far in Nepal is in the range of $ 2,500 per kW. Nepal needs to formulate a policy to choose to implement projects estimated to cost at the reasonable level, especially in view of the upward pressure on retail tariff or negative pressure on NEA’s bottom-line and the country’s macro economy.
Ø Due to the World Bank covenanted restriction on undertaking any project of capacity higher than 10 MW to implement previous incarnation of Arun III, Nepal would have been reeling under load shedding problem from 2000 through 2005. By avoiding such load shedding Nepal succeeded in avoiding adverse impact on NEA, its consumers, GoN and the macro economy of the country. GoN should develop policies for comprehensive options assessment to prudently choose cost effective projects that could be built in time at reasonable cost.
Ø In view of the track record of public sector infrastructure projects, cost overrun is a norm rather than exception. Therefore, the chances of the actual cost per kW of previous incarnation of Arun III increasing from over $ 5,000 to around $ 8,000 level was highly likely and NEA would have been overburdened by such increased cost at commissioning and would have incurred huge losses or would have been impelled to pass such cost through to consumers requiring exorbitant retail tariff. Nepal needs to develop a mechanism to curtail cost overrun.
Ø Further, in view of the track record of public sector infrastructure projects, time overrun is a norm rather than exception. By not implementing previous incarnation of Arun III Nepal succeeded in avoiding prolonged load shedding problem of a higher magnitude. Structure and content of contracts and the implementation thereof, needs to be improved upon to avoid potential cost and time overrun (also in the backdrop of massive cost and time overrun in the case of Middle Marsyangdi). Capability of concerned staff to design and execute/manage construction and supply contracts needs to be improved to preclude time and cost overrun.
Ø Priority should be given for the use of electricity in Nepal which will help industrialize Nepal (which triggers the forward linkaged benefits like employment generation, increase in purchasing power of the people and consequential increase in revenue for GoN) and only excess and spill energy should be exported, especially when electricity generated by a project like this is highly cost effective.
By canceling Arun III in 1995, the then President James Wolfenshon of the World Bank saved Nepal from the vagaries of the failed financial engineering of this project. Besides, because of this action, the magnitude of load shedding now is less severe than it would have been if it was decided to implement the old incarnation of Arun III – load shedding of late 90s would have continued through till now and the years to come and current load shedding would have been of higher magnitude. Further, the consumers would have been burdened with exorbitantly high tariff, due to the high cost per kW of installed capacity with the implementation of Arun III, or NEA would have been incurring losses of higher magnitude. Moreover, the cost on the macro economy due to power deficit too would have been higher. Therefore, this is one positive example of bad financial engineering that got nipped in the bud. But, for the average intelligentsia of Nepal, this is beyond comprehension and, they are going around moaning about the non-implementation of failed financial engineering manifest in the old incarnation of Arun III, even now.
Moreover, the new version of Arun III project, although with double installed capacity and less than half average cost per kW (and therefore requiring lower tariff) will not be contributing to the mitigation of load shedding problem in Nepal substantially as it is being undertaken as a dedicated export oriented project. In sum, the cancellation of earlier incarnation of Arun III didn’t contribute to the load shedding and nor the new version of Arun III is likely to solve electricity crisis when it is commissioned.
 Source: Kantipur of November 13, 2008.
 Mahat, Dr Ram Sharan, 2005: In Defence of Democracy, Adroit Publishers.
 Source: World Bank, 1994: Staff Appraisal Report: Aurn-3 Hydroelectric Project
 Source: Fiscal Year 2007/08 – A Year in Review, Nepal Electricity Authority.
 Source: Fiscal Year 2007/08 – A Year in Review, Nepal Electricity Authority.
 Source: Fiscal Year 2007/08 – A Year in Review, Nepal Electricity Authority.
 Soruce: Annual Report of NEA for 2005.
 Source: www.sari-energy.org
 Source: ADB (2001): Project Completion Report on Seventh Power Project
 Source: http://sjvn.nic.in/
 Shrestha, Ratna Sansar 2008: “The compulsion to live in the darkness even after 10,000 MW”. Gorkhapatra of November 8, 2008.
Published in Hydro Nepal Issue # 4 January 2009