Your “Average Tower” Doesn’t Exist And It’s Costing You Money
1,000 towers. Which 10 get your attention first?
A network average tells you whether the portfolio is moving in the right direction. But it does not necessarily tell you where the money is going. For a TowerCo managing hundreds or thousands of sites, that distinction matters. Grid availability varies from location to location. Grid performance and generator dependency differ. Battery condition, solar contribution, equipment configuration, site load, tenancy and maintenance history all affect the cost of keeping a tower online.
This is why the better management question is not:
“How is the average tower performing?”
It is:
“Which towers are behaving differently from comparable sites — and why?”
That is where benchmarking becomes useful.
Nigeria makes energy performance a site-level problem
Energy is not a secondary operating issue for Nigerian telecom infrastructure.
IHS Towers states plainly:
“Power is our largest single operating expense.”
The company also says diesel's impact is particularly significant in Nigeria because of low grid availability, and that it is investing in hybrid and solar power systems to reduce diesel consumption.
Airtel Africa has described the same operating reality. Speaking about Nigeria, former Airtel Africa CEO Segun Ogunsanya said: “Energy and diesel are one of our cost drivers,” adding that grid availability in Nigeria is very low.
So when a TowerCo looks at a national or regional average, the sites underneath that number matter. A portfolio may appear to be performing within target while a relatively small group of towers is consuming significantly more diesel, running generators longer, experiencing repeated battery events or generating more field interventions than comparable sites. Those sites can disappear inside the average.
Manage the exceptions, not just the average
Consider a 1,000-site portfolio; most towers may be operating within their expected range.
But perhaps 30 sites show unusually high generator runtime. Another 20 repeatedly experience battery-related events. A smaller group consumes substantially more energy than towers with similar loads and operating conditions. The portfolio average may still appear reasonable, but those exceptions may represent some of the strongest opportunities for operational improvement.
This is the principle of management by exception.
Instead of giving every tower the same level of attention, operations teams identify sites whose behaviour falls outside an expected range and investigate those first. The objective is not another dashboard. It is better prioritisation.
Previously, assessing large portfolios took large teams and many hours of work. This can now be done in a relatively short timeframe with AI. This leaves the operations team free to focus on more value-added work.
Compare like with like
Benchmarking only works when the comparison makes operational sense. A multi-tenant urban site in Lagos should not automatically be compared with a rural tower operating with limited grid supply. A site with stable electricity cannot reasonably have the same generator-runtime expectation as a weak-grid or off-grid location.
Useful benchmarking therefore starts by grouping towers according to factors such as:
Grid availability;
Geographic region;
Tenancy and load profile;
Generator configuration;
Battery and solar configuration;
Site type; and
Historical energy demand.
Then the questions become much more valuable:
Which weak-grid towers are running generators longer than other weak-grid sites?
Which sites consume more energy than locations with similar loads?
Which towers repeatedly experience the same power event?
Which site has changed significantly from its own historical baseline?
That is more useful than simply knowing whether average diesel consumption rose or fell across an entire portfolio.
An outlier is not automatically a fault
This distinction is important; a high-cost tower is not necessarily inefficient. Higher consumption may be justified by greater traffic, additional tenants, weak grid availability or a different equipment configuration. The signal that deserves attention is often unexpected behaviour relative to comparable conditions.
Two towers with similar loads and grid availability may show very different generator runtimes.
Why?
It could be switching behaviour.
Battery performance.
Generator sizing.
A change in grid availability.
An equipment issue.
Or simply a change in operating conditions that has not been identified.
An outlier is not automatically a fault. It is a reason to investigate.
Nigerian projects already show why site-level performance matters
The industry is already investing heavily in changing how individual sites consume energy.
At an IHS tower site in Kano, for example, the company converted a site that had previously depended entirely on diesel into a solar-battery-generator hybrid system. IHS reported that solar subsequently supplied approximately 79% of the site's power, while diesel consumption fell by more than 40%. The project also reduced annual CO₂ emissions at the site by 19.2 tonnes. That is one site.
At portfolio scale, the implications become much larger. Under Project Green, IHS has invested in grid connections, battery storage and solar solutions across several markets including Nigeria. In 2023 alone, it reported upgrading 2,750 sites across the programme, reducing diesel consumption by 30.2 million litres and generating $20.2 million in annual power-cost savings.
MTN Nigeria provides another example.
Under its Project Zero programme, MTN Nigeria replaced 86 outdated HVAC units across data centres, switching centres and BTS locations in 2024. MTN reports the programme contributed to lower power consumption and improved diesel efficiency, with savings of approximately ₦508.8 million over two quarters.
The lesson is not simply that solar, batteries, or efficient cooling reduce costs. The more important question for a portfolio manager is: Where should the next intervention happen? That requires visibility into how individual sites are actually performing.
From monitoring to action
Monitoring becomes valuable when it changes what the operations team does next.
A useful TowerCo workflow should move naturally from portfolio level into site-level investigation:
Benchmark → Identify the outlier → Investigate → Prioritise → Act → Measure again.
For example:
Portfolio: Which towers have the highest generator dependency?
Peer group: Which of those towers is performing differently from sites under similar conditions?
Site: When did the behaviour change?
Asset: Is the issue connected to the grid, generator, battery, solar system or load?
Action: Does the site require optimisation, maintenance, resizing, further investigation or CAPEX?
That is the difference between collecting data and using energy intelligence.
Your best towers also deserve attention
Benchmarking should not only identify poor performers. A tower consistently outperforming comparable sites may contain an equally valuable lesson.
Perhaps its generator is better sized. Perhaps battery operation is better matched to the load.
Perhaps its switching strategy is more effective. Perhaps preventative maintenance is working better.
Finding what good looks like gives the TowerCo something that can potentially be repeated elsewhere. So benchmarking should identify exceptions at both ends:
Sites performing worse than expected; and sites showing what better performance looks like.
The question for a TowerCo operations team
IHS currently reports 15,848 sites in Nigeria alone. At that scale, reviewing every tower equally is not practical. The portfolio has to tell the team where to look.
So the question becomes:
If you could only investigate ten towers this week, which ten should they be?
enee.io gives multi-site energy operators a consistent view across grid, generator, battery, solar and site energy performance, helping teams compare locations, identify unusual behaviour and then investigate the assets behind it.
The goal is not more data. It is knowing where operational attention can create the most value.