In recent years it seems increasingly more municipalities, utilities, body corporates, individuals, and in particular landlords are opting for the installation of pre-paid metering systems for both water and electricity in order to exercise control. The reason or purpose of their decision may differ but ultimately the decision to install is always centered on exercising greater control.

I am not in the business to mince my words so let’s just speak frankly here. Municipalities and utilities in particular use the deployment of pre-paid meters to control 3 very specific things:

The 1st thing utility providers consider is Cashflow. When people pay in arrears, they have to wait for their money, sometimes up to 2 months depending on how their billing is structured. Usually, the resellers get payment terms from the likes of ESKOM and the relevant Water Boards. This implies that cashflow typically should not be too much of an issue, provided of course the funds attributed to the cost of sales are ringfenced for this purpose. ESKOM and the Water Boards themselves however have to stockpile coal, chemicals, and other supplies to keep the supply running. They too obtain payment terms from their suppliers but because these suppliers are mostly private sector companies, the terms of supply and expectation of actual payment is typically a bit more onerous.

Unfortunately, a variety of issues including but not limited to mismanagement of funds, high percentages of the population being regarded as indigents (whether they actually are or not) and non-payment by end-users all wreak havoc on municipal finances. The money collected for utilities is either being spent on other essentials (such as municipal salaries) or by virtue of high unbillable usages or bad debts have become insufficient to pay the utility providers’ accounts. These accounts simply fall further and further into arrears ultimately leading to the likes of ESKOM threatening the disconnection of entire towns. Having the option to collect money BEFORE the utility is utilized becomes extremely attractive because it solves (admittedly temporarily so) the immediate cashflow issue. If the pre-pay process is not managed properly though it is a surefire way to land a (badly managed) entity into even deeper trouble.

Secondly, and linked to the first, is the need to control the usage of users that do not pay. We live in a democratic capitalist nation but with significant socialistic tendencies where the state has essentially stepped in to “foster” everyone unable to fend for themselves. This may be due to a variety of reasons such as insufficient jobs, unemployability (due to lack of suitable education or skills), or child-headed households due to aids and other diseases or other reasons. We won’t discuss these here any further other than to say significant pressure has been placed on government to provide for these individuals who for various reasons cannot effectively contribute to the economy. Pressure on municipalities and utilities to provide free services have resulted in bigger than expected losses where some areas have remained unbilled or unmeasured and usage has been unrestricted. This left municipalities and utilities with the need to limit or cap the usage of individuals who are not paying. Legally local government still needs to provide them with some basic services where at least the expectation is that it would be used sparingly because, using the pre-paid metering technology, the “free” allocation may in fact still run out.

Lastly, municipalities have been deploying pre-paid metering technology as a credit control measure where users are known not to pay, not pay in full or not pay on time in a bid to say that if they don’t pay, they won’t be able to use. If a user, historically on a post-paid metering setup, does not pay they are simply forced into the pre-paid system to keep on accessing municipal utilities. This all seems very logical, right?

For individuals opting to install pre-paid for themselves the decision to do so has been informed primarily by two choices. The first is that they wanted control over their consumption and visibility as they use so that they can (inside their household) be more cognisant of what they consume. This functionality can however easily be achieved with a conventional metering solution as well, but it seems some people actually want to be disconnected or have an alarm as a reminder of their usage which conventional metering does not natively do. Full smart conventional meters are however still able to provide this functionality WITHOUT the need to convert from a post-paid to a pre-paid billing setup. Doing the conversion from post-paid to pre-paid metering for most however also gives them the confidence that they will not be hit unexpectedly by an exorbitant bill at the end of an excessively cold winter month.

The second reason individuals have been opting for pre-paid metering is that for usage below a certain amount of units per month it actually works out cheaper to be on pre-paid and that if you consume nothing on pre-paid, you pay “nothing” either. For post-paid, this is typically not true. Whilst the rates are generally cheaper than that of pre-paid, there are some substantial network and availability related charges that get levied regardless of the usage. Even if you don’t use it at all you could still have a substantial utility bill based on the basic availability of the utility to your premises. When one absorbs these charges into your monthly total being spent on utilities post-paid actually works out more expensive than pre-paid base rates for most residential users.

Landlords and Body Corporates have opted to install pre-paid meters mostly to prevent unnecessary losses. Unlike municipalities and utilities, landlords and sectional title and estate body corporates cannot simply turn off utilities when an account remains unpaid, at least not without a court order giving them the right to do so. By installing a pre-paid meter they believe that at a minimum they can “reduce” their potential losses on the rental or levy accounts because at the very least they will not also be paying the utilities on behalf of their delinquent tenants or neighbours. Again, this seems logical on paper even though it is totally wrong. More about this later in this publication.

How pre-paid metering works

In order to really grasp the advantages and shortcomings of the technology we really need to understand how the technology works. Pre-paid metering in its most simplistic form is a utility meter connected to a valve or relay that allows the meter to electronically disconnect a utility if a certain set of conditions are true. These could be any conditions the meter is capable of and programmed to observe.

In most cases, the primary condition is that when the loaded credit on the meter runs out, the utility disconnects by closing a solenoid valve in the case of water or gas or switching a relay in the case of electricity.

Historically the pre-paid meters installed by most municipalities are STS or STS type II meters which adhere to certain pre-paid meter standards. For the purpose of this article, we are focussing mostly on these types of meters as they constitute the majority of pre-paid meters actually deployed in the field. The meters are dumb to the degree that they can mostly receive information but is unable to actively communicate anything back. They are able to receive new tariff tables when the tariffs for utilities in the city change and apply these to the purchase of credits. They can also be programmed to provide a certain amount of utility free of charge at the start of each month or split up on a daily basis across the month.

The manufacturers generally claim that the meters are tamper-proof or tamper-resistant, both of which are not entirely true. The meters can detect when they are simply bypassed or bridged (direct connection between the incoming and outgoing electricity connection of the meter). Some of the “better” models can even detect when the cover over the terminals is removed by a non-technician. What none of them can do however is to detect when a second electrical circuit is powered directly off the supply or when a second supply line in the case of water or gas is tapped off in front of the meter.

This last problem has at least partially been solved with the supply of split meters where the metering unit and the CIU (Customer Interface Unit) from where tokens/credit is loaded has been designed as two separate units. These devices then communicate with each other via PLC (Power Line Communications) or some other communications technology such as Zigbee or other low power radio communications. This meant primarily that the meter could now be installed outside in the street in a kiosk or at the top of a pole feeding the property far away from the grimy paws of those would-be tamperers. Having no direct access to the metering unit and being unable to tap off the supply before it reaches the meter made a big difference in how easy it is to bypass any meter.

The other issue with the concept of “tamper-proof” or “tamper-resistant” pre-paid meters, at the risk of now stating the obvious, is simply that the meters cannot communicate back. If a meter is then tampered with there is no way for the meter to communicate that status back if it is an STS or STS type II meter. So if it cannot communicate its tampered status it can only perform one of two actions… Some meters simply put a frowny face on the display and flashes an error hopefully guilting the offending party into submission and the others actively disconnect the supply. Now, if I was the guy trying to bridge out my meter and the meter actively disconnected the supply as a result of my tampering, why would I not continue and simply bypass the supply entirely? If I then was worried about being caught it is simple to say that the meter disconnected me without warning. Knowing local service delivery and claiming multiple failed calls for assistance after hours, I have bypassed the meter only to not sit in the dark due to the supplier’s “failed technology”. As long as there is credit left on the meter it is extremely difficult to prove which happened first.

The difference between Water & Electricity Metering

There are significant differences in how water, gas and electricity are metered. The true significance of these differences comes in 2 parts.

The first is that whilst these meters measure different utilities they all require electricity to drive their processors and sensors and to effectively do what needs to be done “on-demand”. Since electricity meters are connected to a constant supply of electricity it is not a problem for them. Water and Gas meters cannot easily be permanently connected to electricity. Other than the obvious risks of shock or explosion (should something in fact go wrong) there is also the little logistical problem that there might not always be power available close to where the gas or water meters are installed. In order to overcome this, these meters are almost exclusively manufactured with a battery and circuitry that passes a high level of compliance requirements (such as ATEX certification) in order to be regarded safe, especially around piped flammable gas.

The second major difference is that whilst connecting and disconnecting electricity requires the simple switching of a relay (which needs next to no power to operate), water and gas meters will typically need to close a solenoid valve to disconnect the supply. These valves demand a huge amount of power (when compared to the consumption of a relay) and regular “switching” between open and closed will result in the battery being drained much, much quicker. Of course, these valves are designed to use as little as possible electricity to open or close but it remains relatively power-hungry. The other circuitry controlling the communications, operation, and measurement of the unit also needs to be very carefully considered to reduce the amount of power the meter consumes. Low power radio transmitters, small data packets, and irregular communications with the CIU help to achieve this goal. By lowering these specifications one also has to consider that a high power long-range transmitter is mostly out of the question. The batteries at best are typically made to last a maximum of 5 years in meters that have the ability to disconnect and re-connect supply remotely, whilst some manufacturers still claim falsely to provide meters with “up to” 8 years battery capacity.

Socio-Economic Considerations

Let’s consider the fact that most of these meters are manufactured in first world countries and have been tested under first-world conditions. The average citizen in Europe for instance has a meter like this installed for convenience. He loads utilities every few months as the utilities are relatively cheap in relation to his income and only really gets bothered by the meter when his wife in 10 months time call him to tell him that the utility has run out to which he responds by quickly and remotely loading an additional 10 months worth. The valve probably turns at most twice a year and this results in 10 movements in addition to its measuring, monitoring and reporting functions over a 5 year period. In most first-world countries the utilities are charged at a flat rate per unit which also enables the purchase of bulk utilities over longer periods.

If one were to do the same in South Africa, with our tiered step tariff systems, the utility quickly becomes prohibitively expensive as the vending systems presume that one is purchasing for usage in the current billing period (typically a month). The average African worker (from anywhere on the continent) is not this lucky. Our utility charges in relation to our earnings are astronomical. People generally don’t even buy utilities monthly but tend to load utilities more regularly, some weekly and some even daily. This also results in most people only loading utility credit when they have already been disconnected. Now you have utility meters that disconnect and re-connect supply on a weekly (sometimes even daily) basis. What happens to the poor underpowered battery? It was supposed to last 5 years (under European conditions) but some are running flat in as little as 6 to 12 weeks.

Due to our unique socio-economic circumstance on the continent and a deep socialist stance expecting government and society in general to fend for those who cannot fend for themselves the problem then extends even deeper than this. Let me explain. Previously all utility users were allocated 6kl of water as a free basic allowance. The expectation to continue delivering this free allowance, especially to the poorest of the poor, even after it has been abolished in most districts’ rates systems remain. In many communities very few people are paying for utilities utilised. Recent statistics at the time of writing this article shows that Alexandra in the City of Johannesburg for instance is home to around 50000 families. From these only around 4% are paying something towards their accounts and the remaining 96% then by implication is not. In Soweto around 160 000 pre-paid water meters have been installed with only about 12 000 or 7.5% of users purchasing water credit on a monthly basis. Whilst these figures seem shocking it is in fact true for many poorer and marginalised communities around the country.

As mentioned earlier some municipalities have elected to install pre-paid meters in these communities to limit the level of loss. This meant that after for instance the initial 6kl of water has been dispensed the meter disconnects the supply until the next allocation is due. Unfortunately, this didn’t quite have the desired effect. Communities historically revolted because leaking toilets and other infrastructure on the consumer side of the meter resulted in the entire 6kl leaking out in the first 2 days of the month leaving users without water supply for the remainder of the month. The demand was then made that instead of dispensing 6kl free water per month the meters should be programmed to dispense the first 200l free per day instead. Sounds like a reasonable request right?

The only issue is that no-one partial to these deliberations understood how the meters work or the limitations of the infrastructure. Now you have a meter that was designed to switch on and off maybe 15 to 30 times over its entire lifespan being required to do 30 valve changes in the first month. Meters are connecting and disconnecting utilities daily, running their batteries flat. Once the battery is flat the valve is stuck in either the open or closed position. If it is closed you can bet your bottom dollar that the user connected to it will moan like a drain (no pun intended) until the battery or meter is replaced and their supply restored. If however the valve was stuck in the open position instead you will probably never hear from that consumer again, for anything, as they don’t want to take the chance for any worker to visit and discover their amazing fountain of never-ending water. Regardless of the outcome, the battery in the meter still had to be replaced. Considering the fact that this is meant to conserve maybe R80 of water per household per month, the costs of the battery replacement and the labour to visit to perform the replacement the financial argument for this approach fails very quickly.

Practical & Installation Considerations

In terms of installation and what could and couldn’t work the biggest issue with pre-paid meters is installing the metering unit in a place where the customer won’t have direct access to the meter itself or more importantly to the incoming supply. Bridging a line from the incoming supply will result in the utility going down the bridged line or pipe not being charged for and an under-recovery on the actual usage. At the same time, the customer needs access to a keypad from where tokens can be loaded manually if needed. This then essentially makes an integrated pre-paid meter (where the keypad and metering unit is a single unit) pretty much useless unless you are somehow otherwise ensured that the customer will never be able to bridge out the meter or tap off the incoming supply.  

Let’s focus our attention then on split meters where the metering unit and the CIU (Customer Interface Unit) are two separate units, one installed out of reach and the other plugged in or installed inside the residence. In this configuration it is required for the CIU to be paired with and communicating to the correct metering unit in order to load credit for utilities. This is either done through a separate communication wire, Power Line Communications (PLC) where the comms are running on the same wire supplying the power or in the most modern units with Zigbee radios or similar short-range radio frequency communications.

It is not common to see meters using a separate communication wire anymore as this is older technology which was time-consuming and difficult to install (especially if the meter is in a kiosk in the street or at the top of a pole with the CIU inside the house).

Technical & Communications Considerations

The most commonly used pre-paid metering configuration throughout Africa is PLC but this comes with its own concerns. PLC comms units are usually badly affected by old electrical wiring and older appliances (fridges and the likes) causing “noise” on the power circuits and making communications unreliable at best. In addition, when these meters are installed in close proximity to each other (such as in sectional title complexes) it is like a crowd all speaking at the same time making communications even more unreliable. This is of concern especially to body corporate trustees or landlords who will have to field the calls from residents unable to load credit on their meters for consumption.

More modern pre-paid meters utilise short-range radio technology to communicate with CIU’s. This is slightly more expensive mostly because the meters that deploy these technologies are usually not pure pre-paid meters but instead smart meters which extend the capability of the meter quite extensively. This has mostly been deployed by the likes of the City of Johannesburg to customers still on post-paid conventional credit meters under the ruse of not needing to enter the premises to read the meter. The actual intent had little to do with the reading of the meter although I am sure it did play a considerable role in today’s society of long work hours, lock up and go and lack of trust making entry for meter readers just about impossible. The greater likelihood is in fact their ability to remotely disconnect non-paying customers and in extreme cases even remotely convert them to pre-paid metering without their consent if they are proven to be unreliable payers.

In the case of water metering, the situation is similar although some water pre-paid meters do no deploy a CIU but instead use a proximity tag (RFID or NFC) to load credit purchased (as well as tariff updates etc.) The fact that none of the STS type meters communicates their consumptions (or any other information for that matter) back is also significant. The reality is that in the past the consumption data for municipalities and utilities were used to do a level of network balancing. This meant by monitoring of the bulk meters supplying into an area and the consumption meters on which customers would be billed the maintenance and engineering divisions of municipalities could determine to a reasonable degree where their greatest losses were. They could effectively month on month tell what they were losing in each area and which one required the most attention in infrastructure maintenance and upgrades. Management also utilised this same information to inform budget decisions in order to maintain and replace infrastructure in the different areas they’re responsible for.

The moment even a single pre-paid meter was installed this went out the window. There is a record of how much money was spent and this could be very rudimentarily converted into a rough unit quantity but just because it was bought does not mean it was used. Engineering was officially blinded and now had to almost wait for something to fail before they could say that it needed replacement. Network balancing is no longer an option.

Legal & Outcome Based Considerations

It is right here where things get truly interesting, especially when it comes to landlords, estates and sectional title complexes trying to pass the cost of utilities bought in “bulk” on to their residents. Typically most complexes and estates have a single bulk supply meter from the municipality or utility and separate sub-meters from which each resident is then billed for their individual consumption.

This requires those meters to be accurately read and the consumption translated into an amount which by law may not exceed the cost of the utility to the complex. The complex or estate is forced into billing at the same rates used by the municipality or utility but can realise a slight saving in terms of the steep tariffs and may charge for the reading of meters and other administration charges. The same goes for landlords, REITs, and property managers letting out their property.

In the case of many-body corporates and estates though the increased load on trustees and the hassle in handling queries as well as the prevalence of non-paying owners have convinced them to go the pre-paid route. The thinking in principle is that even though the owner has not been paying their levy account which historically included electricity, by utilising the pre-paid metering system they would be reducing their losses to only the levies which sounds like some solid thinking.

What they do not consider or appreciate however is the fact that whilst previously they could not disconnect or restrict supply to any unit without a court order, they, after deployment of a pre-paid metering system, cannot do so AT ALL.

By deploying a pre-paid metering solution in a complex or estate or even to a single unit as a landlord one has effectively separated the arrears levies or rental account from the utility account. Since the utilities are paid for separately and have been pre-paid, it has no bearing on the arrears levies or rentals, and as such utilities cannot be disconnected, even by court order in order to motivate non-paying residents to pay.

By comparison, it is relatively easy (as case law has now proven) to obtain an order to disconnect or restrict and influence a non-paying resident into paying their levy account. By applying payments received to the arrears levies and other charges first before paying utilities the resident is essentially bullied into keeping their account up to date. This can NEVER be done in a pre-paid metered scenario.

In summation, if your intention is to collect overdue accounts, deploying standard pre-paid meters to limit the losses is a bad idea that will probably come back to bite you.

 

Select deployments for which pre-paid metering could be considered

All is not doom and gloom for pre-paid meters. There are still scenarios in which pre-paid meters may actually be the correct way of dispensing utilities. Some districts have deployed intelligent multi-user pre-paid meters at communal standpipes etc where a user would tag in to open and again to close the supply. This way each user is conditioned into paying for what they use and the standpipes are not left open with running water down the street. Electricity is usually deployed to these central locations but even if it is not replacing the batteries regularly makes sense as the meter services a community instead of just a single family. Driving the concept home that everything costs something to someone is also an important fact of life that everyone needs to remember. 

Pre-paid meters operating on SMART infrastructure with two-way communications also alleviate some of the problems experienced with the all-too-common conventional STS type pre-paid meter and coupled with a reasonable on-site meter audit strategy this may very well solve quite a few problems for the average municipality or utility.

Landlords, body corporates and property managers, on the other hand, should think twice about the deployment of any less than a SMART split meter technology pre-paid setup where other fees are not being collected through the same infrastructure model as it may in fact cut off their (legal) hands when it comes to collecting other mounting arrears. For most of these private entities and individuals the cost implications of implementing such solutions (especially in the residential space) may still be prohibitively expensive.

In reality, as far as the utility dispensing and revenue collection strategies are concerned, there is no one-size-fits-all answer. Over the last 3 decades of municipal revenue protection, collection, and field-work mobility exercises we have found that most government entities require at the very least a combined approach where different meter types are deployed into those areas for which they are best suited and managed by a central system able to collect data from various sources and technologies.