TLTR: In this article I will show you a very unique approach to tower upgrades and foundation retrofits. The methodology helps to reduce significant costs in case a tower or mast capacity have to be increased. The concept is to make a site-specific wind assessment for a tower portfolio in subject and compare each locations' wind speed with the tower's design wind speed it was designed for. In many cases the site specific wind speed will be lower than the later one. In theses cases the tower has extra capacity. These structures can hold more antennas and the tower upgrade and foundation strengthening can be skipped entirely. If you are interested in the details of this approach, please read further or contact us for more info.

### Introduction

In the last 8 years I have designed more than 10k lattice, green-field telecom towers. A minor portion of it, approx. 15% of this amount was eventually built in Africa, the Middle-East and Asia. Just for the curious, the rest 85% are for the so called telecom tower tenders. Sometimes one single tower design set was a base for 120 real-life structures but sometimes only a single piece. So after a small statistics and including my colleagues designs, I should be aware of 50k telecom towers which are up in-the-air. But more importantly what I also know, that the majority of these towers were designed for a basic wind speed provided by the clients. But how those wind speeds were specified? Has the country in question had a proper wind map? Were there any other reliable resource for extreme wind events? Or just a value as 40m/s, because someone heard from someone? We didn't know. Or at least not in the beginning of the mentioned projects.

In reality, those wind speeds from the specs are are higher and they resulting in over-designed towers, wasted steel and concrete materials, thus unnecessary embodied CO2.

So how we can address this problem? The answer is site-specific assessments. In this post I will discuss mainly the basic wind speed as design parameters, but we need to know that there are others too, like topography, exposure, ice, seismic, etc. The TIA-222-H code and even it's predecessor t 1he "G" allows us to carry our site-specific assessment. The TIA-222-H Ch. 2.6.4.1 exactly describes this.

For regions not included in Annex B,  for the special wind or ice regions indicated in Annex B, and for sites where record indicate that in-cloud icing produces significant loads, extreme-value statistical analysis procedures shall be used to establish design values consistent with this Standard from available climatic data accounting for the length of the record, sampling error, averaging time, anemometer height, data quality and terrain exposure.

Further, ASCE 7-10 or 7-16 gives even more details how an Estimation of Basic Wind Speeds from Regional Climatic Data should be carried out.

In the following chapters I'll guide you through a very high level overview about the methodology but focus more on the cost reduction what we can achieve by that.

### Estimate wind speed from climatic data

I keep this chapter short, but highlighting the most important aspects of the procedure. First thing, we need to collect historical wind data from a reliable weather station in the nearby of our site. In reliable I mean that the weather station is up and running at least more than 20 years ago and the last data it sent is not older than 3 days. In general couple of days needed to transit the daily observations from the station to the central servers, making some basic clean up on the dataset and eventually record it in a giant database.

Not all weather stations report wind speeds and wind directions, so in general we have sufficient data from airports or heliports. All around the globe we can found more than 14000 stations like this. These are reporting daily observations according to high quality standards and are also used by very important industries, like the aviation. We, at TowerUp for the WindFinder in fact consider only weather stations with reporting length as minimum of 25 years and last report has been recorded maximum 3 days before the analysis. By this rules we can fulfil the "length of the record" and most probably the "sampling error" requirement by the TIA standard. Once we have all those wind data, we have to make sure that the units of the wind speed and averaging time are also recorded and take into account. There are many more details about the extreme-value analysis procedure which I will not discuss within this article, but for the curious the following are must be part of the procedure:

• Extreme-value analysis method
• Choose probability density functions fit to the dataset (i.e Gumbel, Weibull, Fréchet, GEV, etc.)
• Fitness test of functions
• Sampling errors
• Standard deviation
• Thicker or thinner tails of the PDF
• etc.

After we have collected wind speed data for at least 25 years, cleaned it up and do the math, our probability density functions will look like the following image.

From here we can estimate the 50-year wind speed and also other occurrence, like 100 years, 500 years or 3000 years.

### Consider other factors

To be able to fulfil all the requirements by the TIA and ASCE codes we still have to do some re-factoring. These are the following:

• Height of anemometer
• Altitude factor
• Topography assessment
• Exposure assessment

We are aiming to create a reference base map for wind and that has some requirements. If, for instance our weather station sits 1200m above sea level it means that we are actually measuring a small wind speed compare to what the design codes expect from us to do. The reference basic wind speed should be a 3-second gust wind speed at 10-meter height above the ground in exposure category C.  Once we have all the data from the terrain assessment we can adjust our extreme wind speed and get the reference basic wind speed for the location of the weather station.

If we would be interested in only one single location then it is ok to use the nearest weather station. This can be accomplished by the WindFinder of TowerUp. If not yet, you can sign up for an account here and make some assessment for couple of locations.

In case our aim is to make an assessment on a bigger tower portfolio and practically achieve the promised 60% cost reduction on tower upgrades, then we have to go further.

### Create a regional wind map

Once we have identified all the locations of our sites, then we can create a map for the overview of the sites and also the weather stations in the region. The below is an example with some random sites in Cambodia as an example.

Once we have created our isolines for the wind speeds we can then interpolate easily the site locations and get a very accurate site-specific wind speed for that particular random location. That wind speed will be then compared with the tower or mast on that location was designed for.

Using my experience from the 50k towers which were designed and built in green-fields, I can use some assumptions about the sites. In design stage and even further, in project stage, a tower supplier doesn't really know where the towers will be built exactly. I also have to protect customer data, so in this example I will use generic assumptions which represents well the concept described in this article.

### Summary spreadsheet for cost savings

I'll  create a small spreadsheet with the relevant data to demonstrate an actual savings based on our wind map and the towers with their previously design wind speed. The below table show some tower designs with their associated site-specific wind speed. The below small tower portfolio assumes that the tower owner wants to upgrade them and place more antennas on the structures, thus exposing them for an eventual upgrade. But instead of jumping into the upgrade project, he runs a site-specific assessment for the wind speed to look for potential savings.

Tower
ID
Tower-design
wind
speed
Site-specific
wind
speed
Wind
speed
difference
Original
strengthening
cost
With
TowerUp
approach
1 33 27.1 5.9 $50,000$0
2 33 28.1 4.9 $50,000$0
3 33 26.6 6.4 $50,000$0
4 40 40.5 -0.5 $50,000$50,000
5 33 37.4 -4.4 $50,000$52,000
6 33 38.6 -5.6 $50,000$52,000
7 33 36.7 -3.7 $50,000$52,000
8 45 35.3 9.7 $50,000$0
9 40 35.6 4.4 $50,000$0
10 40 36.0 4.0 $50,000$0
11 40 35.6 4.4 $50,000$0
12 33 33.3 -0.3 $50,000$50,000
13 33 32.9 0.1 $50,000$50,000
14 33 25.0 8.0 $50,000$0
15 45 31.0 14.0 $50,000$0
16 40 31.4 8.6 $50,000$0
17 33 36.6 -3.6 $50,000$50,000
Total $850,000$356,000

What we can really learn from the above table? When we make site-specific wind assessment and we compare the result with the design wind speed the tower was designed for, we can have two outcome.

• One, the site-specific BWS is higher than the tower is designed for
• Two, the site-specific BWS is lower or equal than the tower is designed for

The differences in the table shows both, as it happened that we have a good mix of both. This is however a very likely outcome. This also means that a site-specific wind speed is not only saving cost for strengthening but also rises safety concerns for other structures. Where we have a positive value that means we have a "surplus" in wind speed, meaning the tower is "over-designed". The negative value means the contrary, the tower is actually lacking capacity for that specific wind speed.

We can draw a small conclusion from this table which can be used for high level portfolio management. Tower with around $3 - 5 \frac{m}{s}$ extra capacity of basic wind speed can hold more load, so more antennas. It is difficult to say an exact value in EPA (effective projected area) of antennas, but we can assume something meaningful. We can recall our formulas from the velocity pressure which is:
$$q_z = 0.613K_zK_{zt}KsKeKdV^2$$
It is sufficient to focus on the last variable in this formula, the $V$ which is the basic wind speed. Since its squared it has a significant impact on the wind velocity pressure. This in practise means that the $33\frac{m}{s}$ BWS for a tower causes 28 percent more pressure then the site-specific value $28.1\frac{m}{s}$. This velocity pressure acting not only on the antennas but the tower's structural members too. Since we have less pressure both on the tower and antennas, we actually have extra capacity on the tower as it's members are not fully utilized.
In the above table I used an assumptions that where at least $4\frac{m}{s}$ difference is present, there I consider enough extra capacity for at least 5sqm antenna EPA. That basically mean one more tenant on the tower.

I used a rough estimate for a general tower strengthening as $50.000.-, which includes at least the following costs: • Structural analysis with strengthening recommendation • Design of new steel parts and manufacturing drawings • Manufacturing of new steel with transportation • Strengthening material transportation to site • Strengthening work of the tower • Foundation retrofit • As-built drawings In case you have a more specific figure for the upgrade cost, you may run the math similarly. For sites where upgrade was needed, I just used a slightly higher figure to cover maybe more steel than it was assumed before the site-specific assessment. This figure might seem low, but actually this is only for the increased members of the steel due to the higher pressure. This$2.000.- can cover approx. 1.3t of extra steel.

### Conclusion

We have went through an innovative approach for tower capacity increase, which in practise is basically a load reduction thanks to site-specific wind assessment. I have showed the relevant design standard requirements and also demonstrated a real-world scenario based on own experiences. Finally, it has been demonstrated that a very small tower portfolio can produce a significant amount of savings on a tower upgrade project and the savings can reach \$500k for a tiny 17 towers portfolio.

If you are interested in more details or aiming a tower upgrade program, please get in touch with us for more details.