‘Circular debt to be Rs50-60 billion by July 2020’
BR Research April 8, 2019
An interview with Nadeem Babar, Chairperson Energy Reforms Task Force
Nadeem Babar is a senior executive with extensive worldwide experience in power generation, infrastructure finance and corporate finance. During his career, he has developed, financed and/or managed over 150 power plants of all commercially available technologies, as well as other energy sector assets. He has also been involved in social sector development, especially education.
Mr. Babar started his career as in investment banker in the Project & Structured Finance Group at the investment banking firm of Drexel Burnham Lambert in late 80s. Later, he became a partner in a boutique investment bank focusing on the energy sector. He moved to the corporate sector in 1995 from his last investment banking position at Credit Suisse First Boston where he specialised in International Project Finance. Initially at Coastal Power, and then subsequently at El Paso International after the merger of Coastal Corporation and El Paso Corporation in February 2001, Mr. Babar headed the power business first for Asia, and then globally for these companies.
Mr. Babar was founder and CEO of Orient Power Company (Pvt) Limited until August 2018. He has served on numerous boards in the past. Currently, he serves as the Chairperson of Task Force of Energy Reforms. He is also the Board of Port Qasim Authority, Sarmaya-e-Pakistan and an independent director on the Board of Samba Bank. He is the Chairman of the Board of Progressive Education Network, a section 42 companies that runs 226 schools adopted from the government, with over 48,000 students.
Mr. Babar holds a M.S. in Civil Engineering Management from Stanford University, a B.A. in Economics from Columbia University and a B.S. in Civil Engineering from Columbia University. He was elected to Phi Beta Kappa (Liberal Arts) and Chi Upsilon (Civil Engineering) honor societies for academic excellence.
BR Research sat down with Mr. Babar where he discussed at length the issues faced by the government in the power sector as soon as it took power, the circular debt situation and the way forward. The edited transcripts of the conversation are produced below:
BR Research: Let’s start with what you inherited when the government took charge.
Nadeem Babar: I would first like to explain two major concepts before going to what challenges we faced in the power sector when we came to the office. The first is net hydel profit, and the second is how NEPRA sets tariffs.
The formula for net hydel profit was set 30-40 years ago by AGN Kazi. At that time, there was no private generation, everything was under WAPDA. He coined the term net hydel profit where the idea was to distribute a part of the profit made by WAPDA on hydel generation to the province that provided the hydel source. Unfortunately, as the time passed, there was a general realization that the formula was not only ambiguous and incorrect in some cases, but also not implementable. And eventually the concept of net profit for WAPDA completely changed after new thermal generation capacity started coming online in both the public and the private sector.
KP and Punjab receive a share of net hydel profit, and recently it has been decided that AJK would be getting its share of the profit soon after the completion of Neelum Jhelum Project.
About 10-11 years ago, the federal government stopped releasing funds regularly to the provinces under the argument that the Kazi Formula for net hydel profit is not workable. Roughly about the same time, Council of Common Interests (CCI) was formed after the 18th Amendment, and electricity was shifted to Federal Legislative List Part 2, which essentially meant that all policy related approvals for electricity now rested with the CCI.
The CCI being a representative of the federation as well as the provinces was in constant tussle with the provinces over the disbursement of funds and the workability of the net hydel profit formula. Paying royalty on water became even more complicated with NEPRA in the play as it was part of the generation cost.
Eventually in 2015-16, an agreement was signed between the federal government and the provincial governments of KP and Punjab where KP and Punjab government agreed to the arrears payable to them up until that point; and moving forward until any resolution of the different interpretations of the AGN Kazi Formula was resolved, a rate of Rs1.10 per kilowatt hour was used by NEPRA–irrespective of “profit”.
Keeping in mind that up until the time of the agreement, Punjab had received relatively small amounts, while KP did get some disbursements every now and then, Punjab had a larger share in arrears than KP. Roughly, Rs153 billion was the amount payable in arrears to the provinces where Punjab was owed around Rs82 billion, and KP was owed around Rs71 billion. It was decided that these amounts would be paid in a phased manner in the next 2-3 years.
NEPRA’s tariff determination for the next year is based on DISCOs’ purchase cost in the current year and future commitments in the coming year adjusted for prior year adjustments, which could be positive or negative. At present, the fuel price adjustment takes place every month with about a two-month lag. All other costs – whether it is new plant additions, or exchange movements on fixed cost, or interest rate movements, or taxes etc. – are adjusted quarterly now. In the past, these were adjusted annually.
NEPRA determines tariffs for all the ten DISCOs individually. Historically, NEPRA used to send these tariffs to the government where DISCOs with lower losses had lower tariffs and those with higher losses had higher tariffs; and the government would then announce a uniform tariff across the country for a particular type of customer, which used to be the lowest tariff amongst all DISCOs. The difference between the NEPRA approved tariffs and the tariffs notified by the Ministry was the tariff differential subsidy (TDS), which the government paid to equalize the gap.
More than a year ago, an amendment was made to the NEPRA Act, which everyone is now criticizing and will probably result in further amendments. A new mechanism was introduced where NEPRA could announce uniform tariff. Effectively, what it meant was that rather than the government equalizing the tariff, NEPRA should determine the tariffs, equalize and announce one uniform tariff for a particular slab or customer category. With this amendment, the government lost its right to put surcharge on high-end customers, which it used to offset some of the TDS. This is where we stand today.
With this background, we come to the main problem that this government faced when it took office at the end of August 2018. There was Rs226 billion of prior year adjustments that the previous government had decided not to pass on. Two major portions of that were the net hydel profits, and the capacity payments incurred on CPEC energy projects that had started operating. These prior year adjustments not passed on were based on the indexations that NEPRA had to apply only till the end of 2017; all capacity payments on new additions and the 35 percent exchange movement since the beginning of 2018 up until now had also not been passed on. So, in todays indexed, corrected and passed on cost, we are looking at Rs425billion additional amount that has to be put in the bills.
DISCOs had not been updating their costs to NEPRA for tariff determination for two years due to political pressures, and it was only after the previous government left the office and the interim government took charge that the DISCOs submitted the information that NEPRA had been asking for tariff determination. The existing tariff was at Rs11.71; and in October 2018 – NEPRA notified a tariff of Rs15.31 after incorporating the prior year adjustments of Rs2.18 up until 2017, plus around Rs1.6 for the ongoing year until October 2018. Technically, when the PTI government came into power, the rate should have been around Rs14 had Rs2.18 on account of prior year adjustment having been passed into the rate already. But since it wasn’t, NEPRA proposed a tariff increase of Rs3.82.
There was a lot of debate on this when the matter was taken to the ECC; on one hand, we faced the issue of such an enormous increase in tariffs after only one month of taking charge; on the other hand, the KP government was banking on receiving the net hydel profit disbursement of Rs70+ billion in the current year out of its total development budget, and the CPPA was waiting to pass on the remaining capacity payments for the additional CPEC projects that had come online. After much debate and discussion between the government and NEPRA, the cabinet eventually increased the tariff only by Rs1.27 in January 2019, with further increases to come later in various stages, all to account for past cost not having been passed through.
BRR: What is the deal with the circular debt? Where does it stand?
NB: Let’s start from 2018. In January 2018, circular debt stood somewhere between Rs450-475 billion. The government at that time decided to clear Rs200 billion; between January and May 2018, the then government paid Rs150 billion by borrowing, and another Rs50 billion was sought approval in the last week of the PMLN government, but that never got disbursed. On June 30, 2018, the circular debt stood at Rs600 billion even after all the payments that were made. Why? Because they did not take into account the circular debt being generated during five months of 2018 i.e. Jan-May.
In 2016 and until latter half of 2017, the government was following a policy of revenue-based load shedding, which technically meant that high loss areas had higher load shedding. This kept the circular debt from crossing Rs300-325 billion at any time. In late 2017, the government decided to change this policy and load shedding was reduced to minimal levels ahead of elections; and in the next six months leading up to the elections, there was an additional Rs100 billion circular debts created because of this change in the policy.
The actual creation of circular debt in FY18 that included 11 months of PMLN government and 1 month of interim government was Rs453 billion–about Rs1.3 billion per day. Essentially, three extraordinary elements contributed gigantically to this sum. One, other than the TDS, there was Rs175 billion worth of unfunded subsidy i.e. announced by the government, passed onto the customer, but never paid to the distribution companies. This included no allocations made for the subsidy that was announced and was being paid under the industrial support program and the agri support program for cheaper electricity. Also the Balochistan tube wells subsidy and AJK subsidy remained unfunded.
The second element was changing the revenue-based load shedding policy. Third was not incorporating and passing on the actual cost that the government had agreed with the provinces as NHP and paid to new projects as capacity payment. The fourth element, which wasn’t new, was the continued losses that added to the circular debt stock. On a consolidated basis, AT&C losses were around 24 percent for 2016-17. These increased to 29 percent for 2017-18 – the highest ever. This is all what the new government had to contend with when it took charge.
BRR: Where does it stand now given that you have been in power for the last 6-7 months?
NB: The rate increase of Rs1.27 came into effect on January 1, 2019 after a couple of months of deliberation, debate, and review. Government told NEPRA that the tariffs cannot go up by Rs3.82 in one go. What was essentially done was that some tariff rationalisation where the increase proposed was reduced from Rs3.82 by about 50 paisa and NEPRA was asked to spread it over a two-year period – Rs1.27 of which has already been jacked up in January. Let me remind you that out of the remaining Rs2.15., there is a one-time element in the shape of prior year adjustments, which will drop off once it is recovered over the two-year time period.
Coming to your question now; the consequence of all this was a higher run rate of circular debt for the 1HFY19 as the tariff increase came into effect only in January 2019, which translates roughly into a circular debt of Rs220-230 billion for the first six months which is the same run rate that was continuing in 2017-8. However, this number will come down to Rs100-110 billion for 2HFY19, which means that the circular debt being created for the ongoing fiscal year, 2018-19 will stand around Rs320-330 i.e. Rs100 billion lower than what it was in 2017-18.
In 2019-20, two more rate increases would have to come into effect as per NEPRA, and our estimates show that the circular debt for the year 2019-20 would drop to Rs110 billion. And when the following fiscal year starts in July 2020, our estimates suggest that the circular debt will be down to Rs50-60 billion for the whole year, which would be manageable.
BRR: What policy is the government adopting for load management now?
NB: We have gone back to the previous regime where any feeder with AT&C losses of less than 20 percent will not have any load shedding. Around 50-60 percent of the feeders in the 10 DISCOs have no load shedding policy because their AT&C losses are below 20 percent. We have four categories of losses for load management: 20, 30, 40 and above 60 percent. On average, total revenue based load shedding, based on this policy in the country is around 2500MW. However, wherever the AT&C losses drop down, the revenue based load shedding is reduced accordingly.
BRR: Can you tell us a bit more about the unfunded subsidies? What is government doing to address this problem?
NB: In November 2018, when all this was being debated in the Cabinet, I made an impassioned plea to the Prime Minister and the Finance Minister to stop the practice of unfunded subsidies if things are to be fixed. They both readily agreed. The Finance Minister was right in saying that he had no budgetary room for more subsidies and has committed to pay the subsidies for zero-rated export sector, which amounts to Rs24 billion.
Programs like agri support where electricity is being charged at Rs5.35, and industrial support program where industries are getting a discount of Rs3 are continuing as it is difficult to finish them. However, the government was able to solve the AJK issue in the Cabinet where rate has been increased to Rs5.79 versus the previous rate of Rs2.73. The ECC has approved it and it will be in effect soon.
A large part of the Balochistan tube well subsidy that has both the provincial as well as a federal component has not been paid for the past 10 years. This subsidy is the single largest component or around Rs197 billion in the Rs950 billion receivables of the DISCOs from all types of customers. Ten years ago, the estimated cost of a tube well at that time was calculated to be Rs75,000. And agreement was reached back then that the farmer will pay Rs10,000 of the cost, while the federal government and the Balochistan government will pay 40 and 60 percent of the remaining Rs65,000. In case the cost was above Rs75,000, the excess was to be paid by the farmer. To address this situation for the future, the government has planned to solarise all these tube wells in Balochistan. This will take at least two years, and we have found international donors as well as commercial companies that are interested.
BRR: If you could summarize, what are the key focus areas for the government today?
NB: Against this backdrop, I identified a few areas for the government in late September last year. After several rounds of meetings with the PM, ECC, CCoE we agreed to follow these areas, which should bring down the run rate in the circular debt to Rs100-150 million a day by July 2020.
First, there will be no unfunded subsidy in the upcoming budget. Second, DISCOs must recover the prior year actual costs. Almost 50 percent of these past costs have to go to the two provinces in shape of net hydel profit, while the remaining 50 percent is the additional capacity payment that remains unrecovered. As I mentioned earlier, there is a time horizon for this. In case we don’t drop off once these costs are recovered and keep them part of the tariff for some time we can actually generate money to retire all the circular debt that was parked on the balance sheet of Power Holding Company earlier. If we drop these items off, we will have to deal with the circular debt separately through a longer program.
Third is the theft control program that was launched mid-November 2018 in Punjab, and mid-December in Sindh and KP. Effectively, we have 2.5-3 months’ worth of data, which shows that we have been able to increase revenue by Rs42 billion in this time, out of which Rs20 billion has come from higher units sold and hence higher rates charged since January 2019, and Rs22 billion has come purely from reducing AT&C losses. The AT&C losses have come down by 1.8 percent in 2.5 months. Our target is to bring the AT&C losses down by 3-4 percent to 25-26 percent by the end of June 2019. The target for next year i.e. FY20 is to bring it down to 22 percent, which would essentially mean that we would have contained the leakage to about 6 percent given that NEPRA’s threshold for AT&C losses is 16 percent.
The theft control program has two parts: law enforcement and technology intervention. Law enforcement is in full swing; there are over 20,000 FIRs registered against electricity theft; more than 2000 customers are in jail; more than 500 government personnel are behind bars. On the technology side, we are trying our best to reduce manual intervention to minimal.
Coming to the last item; we have to reduce our generation cost. Our average, generation cost is high and 42 percent of it is based on imported fuel. In the last one year, the cost of generation has gone up by Rs1.5 per unit only from the 35 percent currency depreciation during the year. Our average weighted basket cost of generation at all the plant gates in January 2018 was about Rs10.5. And the actual T&D costs were around Rs2.10-Rs2.15, which included the distribution and transmission staff costs, O&M costs, etc. So, the true cost of delivery without taking into account the losses was around Rs12.7. After including the 16 percent allowed losses by NEPRA, the true cost of delivery would mean a cost of over Rs14.5. Adding the 13 percent disallowed losses – 29 percent total losses minus 16 percent allowed losses – the figure would be over Rs16. So there was a big gap between this figure and the weighted average rate of Rs11.71 at that time, which was piling up and going into circular debt.
Also, we have to realise that we cannot just look at the present day analyses; we have to look at the future implications over the next five years. One of the first things I told the government was that we need to plan our generation on the basis of a 25-30 year forecast preferably month by month. The PM was very supportive; so I launched something I called Pakistan100, which for me is a 28-year plan for power; I will follow this for the petroleum side and then integrate the two. The first draft for power has been submitted. The idea behind this analytical exercise is to get out of these alternating cycles of shortages and excesses; start planning at the generation level and match it with transmission and distribution planning; redirect our fuel mix by policy; shutdown a good chunk of power plants from the existing fleet that are too inefficient; and change the forex dependability.
BRR: We’ve been hearing about smart metering for some time now? Is the government doing anything in this regard?
NB: Yes. To curtail losses, we are launching the AMI Program, which is the smart metering system. We have received the first tender for IESCO, funded by ADB. This month we are launching for LESCO. And by the end of the year, we will be floating the tenders for PESCO, HESCO and SEPCO. AMI or Advanced metering infrastructure is an integrated system of smart meters, communications networks, and data management systems that enables two-way communication between utilities and customers. There has been a lot of controversy around this idea in the past, but that has been addressed, and I believe that the program should be effective.
BRR: When do you expect to see smart meters at the household level?
NB: Converting all meters to smart meters across the country is a five-year program. If you talk about IESCO, the contract will be awarded in June this year and the program is for 2 years. LESCO contract is expected to be awarded in July 2019, and it will also take around two years.
BRR: One issue raised on the AMI is the local meter manufacturers don’t qualify due to some conditions in the ADB contract, which technically means that the equipment will be imported. They argue that this would affect local business and job creation that could get a boost otherwise. What’s your take on this?
NB: This is half true. They have approached the High Court in Islamabad. The tender has some technical and some financial conditions for you to bid. Initially when the program was conceived some 4 years ago, technical conditions proposed that if you are a sole bidder, you need to have the experience to deploy these kinds of devices in at least 3 locations globally, which these local players didn’t have. The current tender says that if you are part of a consortium, at least one member of that consortium has to have experience in 3 locations. It further says that if you want to be a named subcontractor, you need experience in one country, which could be your own. The local players technically qualify the subcontractor condition. So local players can meet these conditions either as part of a consortium or a subcontractor.
On the financial side, there are two triggers. The first one is that the annual turnover of the single bidder or the consortium should be $80 million for last three years. Second condition is that if you are part of a consortium, then as a member of the consortium to have to have an annual turnover of $20 million for last three years. According to these local players, none of them qualify the $20 million annual turnover condition. It is my belief that at least PEL qualifies.
BRR: Coming to generation, what are your plans and targets going forward?
NB: Our current name plate capacity is 33,000MW, while 26,000MW is the dependable capacity. By 2030, we plan to increase generation capacity to 52,000MW; and the imported fuel component of generation that stands at 42 percent today, we plan to bring it down to 20 percent by 2030.
In the next 5-10 years, our main focus will be on renewable energy; the hydel plants that are in process – committed and identified; Thar coal plants that are awarded, in construction or in various stages of contracting; and finally three nuclear plants under construction. The policy is strictly no new imported coal plant; no new imported LNG plant. By 2030, hydel and renewables would account for 65 percent.
BRR: Renewables have a criticism that they cannot become the base load because of fluctuation in power generation as well as the demand in that area. But your targets seem to show that at 65 percent, renewables will become the base load. How do you address this query?
NB: We are bringing a new renewable policy, which will address these issues. The new policy has a radically different approach. The cost of renewables at this time is in most cases lower than the fuel cost of oil, and in some cases RLNG. If inducting renewables lowers our average cost of generation, then we are not worried about additional capacity because it improves our financials. It is correct that variable renewable energy (solar and wind) causes problems of voltage fluctuation for the weak grid that we have.
How we plan to counter this is by setting a target for renewable share and identifying what we need to do for the grid to absorb them at the least cost without any variations. We have just signed up with the World Bank for a study of the entire country on resource mapping together with interconnection mapping, which they have promised to finish in 6 months. There is another study being conducted in parallel by Lahmeyer, Germany funded by the World Bank to identify what points in the main grid need strengthening if we plan to increase renewables to 20-25 percent in the system. Based on the results of these studies, our new policy includes capacity auctions every year for solar, wind and biomass.
BRR: What is stopping you from commercialising solar panels in the residential and commercial sector especially when we need to increase indigenous generation?
NB: Nothing is stopping us. If the new renewable policy is approved as is, it has an entire section on this issue. I have proposed to completely open up this segment: billing, net-metering, wheeling and excess being sold back. This will attract resistance though. DISCOs have a legitimate concern that their better paying clients will start to shift and we have to address this concern through market forces.
BRR: Coming to furnace oil, what is the way forward?
NB: Back in November, when the refineries threatened to shut down, I was called in to sit with them. We have had three rounds of negotiations with all the refineries and we reached an understanding on a long term program. We have worked out a two staged program. In the first phase, the refineries need to have their export channels developed as soon as this summer is over.
We are going to add some minor infrastructure at the port to have reversed pumping and reverse decanting to fill a ship with FO for export. PSO has been instructed to remodel the storage capacity to increase the pumping capacity by the end of this summer to be able to ship out a cargo of 50-60K tons every 10 days. Our production of FO is about 300,000 tons a month.
The next step is that like the new refinery policy with tax concessions, we are going to give a 5-year window to the existing refineries with similar concessions to upgrade. We have also suggested that rather than every refinery having its own hydrocracking unit, they should all have a single unit in the country that should do the job. To recap, refineries will export furnace oil for the next two years where we will give them some financial incentive in the first six months as FO is a negative margins product due to import price parity. After that, they have to compete on their own or go to deeper conversions.