Interview: ‘India is losing Rs 100,000 crore in unbilled electricity. The solution is smart meters’

In a bid to improve power utilities’ financial health by increasing their revenues through efficient billing, India plans to replace the existing 250 million traditional electricity meters with smart meters that use digital technology to enable a two-way flow of electricity and information. Traditional meters only record energy consumption for billing purposes.

Smart meters are crucial for reducing electricity distribution companies’, or discoms’, losses. In the financial year 2018-’19, these stood at Rs 27,000 crore, four-times the allocation for the renewables ministry in 2020. They do this by not only improving billing and revenue collection but also reducing the difference between the cost of supply and the revenue collected. A smart, automated metering system, without manual intervention, would reduce meter-reading and data-entry errors and costs.

These meters would also estimate consumer demand, letting utilities forecast and contract for power requirements more accurately. This is essential for integrating renewable energy into the grid using the “time of the day” policy, as we further explain.

To smoothen the transition and create an ecosystem for smart meter use, the Central Government, in 2017, created the Energy Efficiency Services Ltd or EESL, a joint venture between several public-sector enterprises helmed by the power ministry. EESL’s best-known success has been speeding up the adoption of fuel-efficient LED lights by bringing prices down by 80%. The company floated its first tender for smart meters in August 2017. On February 24, it announced that it has successfully installed one million smart meters in the country.

States that have installed smart meters include Uttar Pradesh, Haryana, Bihar and Delhi. Some discoms using smart meters have increased their per-meter, per-month revenue by Rs 200, said Saurabh Kumar, managing director at EESL. This increase in discoms’ revenue is mostly due to improved monitoring efficiency of per-unit electricity supplied and improved billing due to smart meters, he added.

Kumar has worked with the government in various capacities. At the Bureau of Energy Efficiency, he led the implementation of the world’s largest Clean Development Mechanism-efficient lighting project to reduce carbon emissions. An electrical engineer from the Indian Institute of Technology, Kanpur, Kumar has a postgraduate degree in public policy from the National Graduate Institute for Policy Studies, Tokyo, Japan.

Edited excerpts from the interview:

How many smart meters has EESL installed so far?
So far, we have successfully installed 1.10 million smart meters. We have completed a project in the New Delhi Municipal Corporation area about a year ago, with some 55,000 smart meters installed. Our projects are on in Uttar Pradesh, with all five distribution companies in the state; in Haryana, we are working with both the distribution companies. We have just started in Bihar.

Are these smart meters showing results? How are they improving discom revenues?
With these 1.10 million meters, the average increase in revenue per month per meter is Rs 200 [when the national average bill is about Rs 450, assuming average consumption of 90 units per month at Rs 5 per unit]. The highest increase was recorded for NDMC, around Rs 500, and the lowest would be in Kanpur and Meerut, about Rs 130. What we are charging to these discoms is Rs 85 for a single-phase meter and about Rs 105 for a three-phase meter.

Numbers are clearly showing that discoms’ revenue is up and there are two main reasons. The first is the elimination of incidences of suppressed demand load. Let’s understand this with a very simple example: Let’s say you took an electricity connection five years ago when you had one air conditioner in your house, and we all know that there is a two-part tariff [i.e. if a discom gives you a 2.5-kilowatt sanctioned load for your house, you will pay some demand charge for it]. Now, over the next five years, you added three more ACs, which means your actual load has gone up from 2.5 kilowatts to about seven kilowatts. But there is no way, in the manual system, to check this. The moment we installed smart meters and put the data online, the discom knew the exact load used in a premise. Therefore, the fixed charges for sanctioned loads go up, increasing the revenue. In fact, [New Delhi Municipal Corporation] and Uttar Pradesh are the biggest benefactors of this.

The second reason is very crucial because a lot of discoms are still saying that they do not need smart meters because their billing efficiency is 99%. [New Delhi Municipal Corporation’s] billing efficiency is 99.8% today and its average bill has gone up by 25%. The reason is, in the manual system you may be billing 99% of consumers, but there is no way of ensuring that the quality of billing is good. There are hundreds of dysfunctional meters and discoms continue to bill at flat rates. Just because you are generating a bill does not mean that you are getting all the revenue that you need to bill for.

The third benefit of a smart meter is transparency. About 99% of consumers in UP who have smart meters are now paying their bills on time–urban and rural. This is because now they know exactly what their consumption is, they do not have to wait for the bill to come after three months. We are now encouraging everyone to download a mobile app through which they can see their daily electricity consumption. Because of this transparency, disputes over bills have also gone down dramatically.

In Bihar, we are taking this one step further. We are going for a prepaid model there. You pay upfront for the energy you would like to consume in the near future [which helps consumers plan and regulate their energy use and bills]. If this happens, imagine how much it will benefit the discoms –they will need no working capital because they get the money upfront before you buy electricity. The need for a huge working capital is exactly the problem they are struggling with today. I would say the exercise has been an enormous success.

How will these smart meters improve the electricity sector in general? How can renewables benefit from them?
Let me first start by telling you why smart meters are a critical element in the future health of the sector. The average billing efficiency in India is 83%, which means that 17% of the electricity in the country is not billed for, forget about revenue collection. This 17% means a staggering loss of Rs 100,000 crore [equivalent to India’s education budget for 2020], if I calculate the per-unit cost at Rs 5. There is only one solution to plug this: smart meters. We have seen smart meters solving the problem in UP, where Aggregate Technical and Commercial losses [the difference between the total electricity units supplied by a discom and the total units billed for] have gone down by 36% in certain feeders [transmission lines that take electricity from the spot of generation to distribution points].

Now coming to renewables, what is the problem today? Why are states reluctant to take renewables? Because renewable power generation mostly happens during the time [of the day] when I don’t need it, not even with the per-unit cost of Rs 2.44 [because there is not enough demand given Indian usage patterns]. The solution to this is the “time of the day” policy [under which different rates apply at different times of the day. Usually, under this policy, the rate of the electricity is kept cheaper during off-peak demand hours to compel users to shift most of their non-essential power usage in that time-frame. This helps a user save money along with helping a discom reduce pressure on its grid during peak hours.]

Now, how do you implement “time of the day” policy? You tell a consumer that I will give you electricity at a cheaper price of Rs 3 per unit between 11am and 4pm, so shift your non-essential usage to these hours. But this can only be done with smart meters because it provides you with real-time flexibility. The other benefit of smart meters is incentivising people by demand response, which means that everyone wants to switch on their AC at 6pm...If I tell a consumer to switch on the AC at 9pm and he [or] she will get a rebate of Rs 1 per unit for those three hours, the load on the grid will shift. This can only happen in real-time with smart meters.

What are the challenges that Energy Efficiency Services Ltd faced with smart meters?
The first major challenge we faced had several ramifications. It was about the integration of the data that is coming out of smart meters to the head-end system [hardware and software that receives the stream of meter data] to the metre data management system [which helps process the received meter data for insights] and then the legacy software of billing and collection. But I am very proud to say that we overcame that challenge.

Now we have been able to standardise the process up to the meter data management system. So now when we start working with a new discom, the only challenge we face is integrating the entire system to their legacy billing software because all discoms use unique billing software. The other challenge we faced was the supply of smart meters in such huge numbers. The supply, however, has now improved considerably. At the moment there are six- [to] seven-metre manufacturers who have got certification from the Bureau of Indian Standards [the national standards body under the Ministry of Consumer Affairs]. We cannot install any metre that is not certified by the [bureau]. In fact, the minister and secretary of power have been meeting with metre manufacturers on a constant basis and we are seeing the results. The number of people on board is increasing by the day.

So is EESL trying to figure out a single solution that works with all the legacy billing software of discoms? Or is it using a different integration exercise for every discom?
It all depends on what softwares the discoms are using. Up to the meter data management system, we have achieved standardisation because it is either us or our partners who are controlling the infrastructure. The only thing that belongs to the discoms is their own billing software. For example, in [Uttar Pradesh], the five discoms have been using three platforms. We are trying to make it common. In the end, UP’s discoms would like to see a consolidated dashboard with all the data about electricity consumption and trends and that is what our endeavour is.

We are bringing all the meter data on a common platform and further, we are planning to have different buckets for different discoms in a virtual framework because all of it is cloud-based. And to achieve this exact model took us a while because no one has ever done such an exercise at such a massive scale in India.

Why GPRS, why not other advanced technologies like radio frequency mesh that experts believe is better, as IndiaSpend reported? [Radio frequency mesh technology uses radio waves to communicate among groups of meters that send data to a data concentrator unit with a SIM card which then forwards it through the telecom network to the discom’s main server. This is unlike the GPRS technology where each smart meter needs its own individual SIM card to communicate with the server.]

We are going with GPRS and possibly looking at the [Narrow Band-Internet of Things], which is a specialised band only for machines on a GPRS kind of a network. [Radio frequency] is a good technology but it does not suit our business model because it is capex heavy [requires capital expenditure to build a new network] due to which our cost-per-meter would go up to Rs 6,000 per meter. It is Rs 2,500 in the GPRS system.

Also, with RF, I will have to build institutional capacity to maintain that network. The benefits of GPRS technology is that I am getting data cost at a very reasonable price of Rs 4 per SIM card per month. And the availability of the GPRS network is 98.5% in the country.

I am not saying that moving forward we may not look at other technologies. We may. For example, in multi-storeyed buildings you can have an [radio frequency]-technology network because the data from all the apartment meters will come to a single DCU and from there go to the GPRS network. So we may look at a combination of network technologies going ahead.

Telecom companies are moving beyond 2G and 3G. Will that not cause a problem?
Our first tender was for 3G network technology. Let us look at what is in it for a telecom operator. An operator is currently getting only 1.1 million consumers for data, whose bandwidth requirement is so small, probably comparable to a credit card swiping machine, that an operator would never want to discard it. Secondly, for an operator these consumers are for life. EESL may get out of the scene but that SIM card will remain in a consumer’s house forever.

Is connectivity not a problem for rural areas?
As I said before, all we need is a telephone line because the data requirement of these meters is very low. And we have worked in rural areas of Haryana and Uttar Pradesh. So far, we did not face any issue with network connectivity. In Bihar, the network coverage is about 99%. In rural Uttar Pradesh, it is almost the same.

What do the next couple of years look like for smart meters?
The government has come out with a very ambitious plan of replacing all 250 million meters in the country with smart meters in the next three years. For this, the government has spoken to the entire meter industry to gear up for this drive. The good part is that most meter manufacturers said that they only need three-six months’ time to be ready to deliver 60 million [to] 80 million meters every year.

All the discoms that we are working with understand that they do not have the internal capacity to manage a project like this. They have tried to issue tenders to install smart meters on their own, but failed. But now they are seeing the results. We are also taking steps to build capacity within these discoms so that even after we leave there will still be people working there who understand this technology and how to take it forward.

Also, in our past experience of procuring meters and system integrators, we have identified a very large opportunity in the sector going forward. So last November we tied up with the National Investment and Infrastructure Fund, which manages the sovereign wealth fund of India, to establish a special purpose entity [a subsidiary], the Intellismart Infrastructure Private Limited, only for the purpose of implementing the smart meter project. The vision behind this move is that Intellismart will be the biggest system integrator in the country which will provide high-quality service in the transition period at a very affordable rate.

How is EESL preparing itself to ramp up meter installation?
Now, we are at a stage where it is simply a question of how many meters you can install per day. Today, I think we are at a rate of about 8,000 meters per day and the plan is to ramp it up to 15,000 to 20,000 in the next two-three months. By the end of this calendar year, we want to reach 70,000-100,000 meter installations per day. Currently, we have UP installing four million meters with us, as per the existing agreement. With Haryana, the agreement is of one million meters, but there are another one million consumers for which we are already in discussion with the state. In Bihar, we are currently installing about 1.8 million meters.

As for our new projects, we have signed an agreement with Port Blair. In Rajasthan, we have actually won a competitive bid to install 500,000 smart meters. We are also in conversations with Arunachal Pradesh, West Bengal and Telangana for hundreds of thousands of meters.

This article first appeared on IndiaSpend, a data-driven and public-interest journalism non-profit.