Climate Considerations in Silo Design: Wind, Seismic & Thermal Loads That Matter

The Night the Wind Almost Took Down a Silo in the Philippines

Here's the thing about climate loads. They don't care about your schedule. They don't care about your budget. And they definitely don't care about the assumptions you made in the design office. I remember a project in Cebu, 2019. We were commissioning a set of four bolted steel silos — 18 meters tall, 8-meter diameter, storing rice hulls. Standard design, or so we thought. The structural engineer had specified 3mm shell plates for the upper courses, 4mm for the lower, based on internal pressure and dead load. Wind load? He'd used the basic formula from the local code. Seismic? The Philippines sits on the Ring of Fire, so we had seismic provisions. Thermal? Cebu's tropical. Everyone figured temperature wasn't a factor.
Typhoon Kammuri hit three weeks after we finished erection. Peak gusts recorded at the site: 165 km/h. The silos survived. Barely. Two of the four developed visible buckling in the upper shell courses — not enough to fail, but enough that we had to reinforce them before the client would accept the project. The failure mode wasn't what anyone expected. It wasn't direct wind pressure on the windward face. It was vortex shedding on the leeward side, creating alternating low-pressure zones that pulled the thin shell inward. The 3mm plates on course 5 and 6 had a buckling resistance of roughly 4.2 kN/m² under external pressure. The actual dynamic pressure differential hit 5.8 kN/m². That's when I learned that climate considerations in silo design aren't just check-box items. They're the difference between a structure that stands for 30 years and one that needs emergency bracing before you've even handed over the keys.

Why Static Wind Analysis Isn't Enough for Tall Silos

Most engineers — especially those who don't work in bulk storage — think wind load is straightforward. You take the basic wind speed, apply a pressure coefficient, multiply by the exposed area, done. Wrong. Or rather, incomplete. For silos, wind does three distinct things:

• Direct pressure — the obvious one. Windward face gets pushed. Leeward face gets suction. This is what most codes calculate first.
• Vortex shedding — when wind flows around a cylindrical structure, it creates alternating vortices downstream. For silos between 15-35 meters tall with diameters of 5-12 meters, the Strouhal number (typically 0.2 for cylinders) puts the shedding frequency right in the range that can excite natural structural frequencies.
• Lift forces — the curved surface creates asymmetric pressure distribution that can generate significant lateral and even uplift forces.

Per ASCE 7-22, the velocity pressure exposure coefficient (Kz) varies dramatically with height and terrain. In Exposure C (open terrain), at 20 meters, Kz is approximately 0.90. At 10 meters, it drops to 0.62. That's a 45% increase in wind pressure just from height — and most silos are taller than 10 meters. Here's a calculation I ran on that Cebu project after the typhoon. Basic wind speed: 70 m/s (per PAGASA data for that region). For a 3mm shell course at the top of the silo:

Dynamic pressure q = 0.6 × V² = 0.6 × 70² = 2,940 Pa = 2.94 kN/m²

With external pressure coefficient Cp = -1.4 (leeward/suction) and internal coefficient Cpi = +0.2:

Net external pressure = 2.94 × (1.4 + 0.2) = 4.70 kN/m²

Buckling resistance of 3mm steel shell (r = 4000mm, E = 200 GPa): ~4.2 kN/m²

See the problem? The net external suction exceeded the shell's buckling capacity by 12%. That's how you get dents. The fix was straightforward — we increased the upper three courses to 4mm and added internal stiffener rings at 1.5-meter spacing. Cost? About $2,800 per silo. Cheap insurance compared to a catastrophic failure. As discussed in our guide to steel silo shell plate design, the thickness-to-diameter ratio is everything when external loads enter the picture.

Seismic Load Calculations: Where the Math Gets Real

Let's look at a real seismic calculation. Project: a 25-meter-tall welded steel silo, 10-meter diameter, storing wheat. Located in seismic Design Category D (moderate-to-high seismicity). Site Class D (stiff soil). Key parameters:

• SDS (short-period design spectral acceleration): 1.0g
• SD1 (1-second spectral acceleration): 0.6g
• R (response modification factor for silos): 3.0 (per ASCE 7 Table 12.2-1, "Bins and hoppers")
• Effective weight W: approximately 450 tonnes (structure + stored grain at 80% fill)

The seismic response coefficient:

Cs = SDS / (R / Ie) = 1.0 / (3.0 / 1.25) = 0.417

But capped at Cs_max = SD1 / [T × (R / Ie)]

For this silo, fundamental period T ≈ 0.4 seconds (short, stiff structure)

Cs_max = 0.6 / [0.4 × 2.4] = 0.625

Governing Cs = 0.417 (since 0.417 < 0.625)

Base shear: V = 0.417 × 4500 kN = 1,877 kN That's nearly 1,900 kN of lateral force at the foundation level. For a structure weighing 4,500 kN total. Think about that — the earthquake is trying to push the silo sideways with a force equal to 42% of its total weight. Now here's where it gets interesting for silos specifically. Unlike a building, the mass distribution isn't uniform. The stored grain creates a hydrostatic-like pressure profile that's highest near the bottom but contributes significantly to the dynamic mass at mid-height. The impulsive mass (material that moves with the wall) might be 60-70% of total fill weight. The convective mass (sloshing grain near the top surface) adds another dynamic component. Most engineers oversimplify this by treating the grain as rigid mass attached to the structure. That works for preliminary sizing. For final design in high-seismic zones, you really should use the impulsive-convective model from ASCE 7 Chapter 15. Foundation design gets brutal in these cases. The overturning moment on our 25-meter silo was approximately 28,000 kN·m. With a 10-meter diameter base, that translates to a maximum bearing pressure variation of ±360 kN/m². If your foundation bearing capacity is 200 kN/m² (typical for stiff clay), you need a mat foundation or deep piling — no getting around it.

Thermal Expansion and Contraction: The Hidden Hoop Stress

Everybody forgets about thermal loads. I've reviewed designs from engineers who nailed wind, nailed seismic, and completely ignored the fact that their silo sits in a desert where surface temperatures swing 50°C between night and day. Here's the physics. Steel has a thermal expansion coefficient of 12 × 10⁻⁶ /°C. For a silo with a 10-meter diameter:

Diameter change per °C: ΔD = α × D × ΔT = 12×10⁻⁶ × 10,000 × 1 = 0.12 mm/°C

For a 50°C daily swing: ΔD = 0.12 × 50 = 6.0 mm

Resulting hoop strain: ε = ΔD / D = 6.0 / 10,000 = 600 × 10⁻⁶

Hoop stress: σ = E × ε = 200,000 × 600×10⁻⁶ = 120 MPa

120 MPa. On a silo that's also carrying internal pressure, dead load, and possibly wind or seismic loads. That's not trivial — it's roughly 40-50% of the yield strength of S235 steel. In hot climates, I've seen the opposite problem too. A silo filled with hot grain (say, freshly dried at 60°C) in an ambient environment of 5°C creates a massive temperature gradient through the shell wall. The inner surface expands, the outer surface stays cool. This generates compressive stress on the outside of the shell — a perfect setup for elastic buckling. The material selection process needs to account for this. High-temperature environments might push you toward aluminum for certain applications, or require stress-relief considerations in welds. Cold-climate silos have their own headaches. Concrete silos in Scandinavian or Canadian installations deal with freeze-thaw cycling that degrades the shell surface. Steel silos in those same climates experience contraction stresses that accumulate over thousands of thermal cycles. Expansion joints — yes, on a silo — become necessary above certain height thresholds.

Load Combination Analysis: Where Climate Forces Collide

Here's the part most people get wrong. They design for wind. They design for seismic. They design for thermal. Then they forget that nature doesn't ask permission to combine them. ASCE 7-22 Section 2.3 gives us the load combinations. The relevant ones for silos with climate loads:

• LC1: 1.4D
• LC2: 1.2D + 1.6L + 0.5(Lr or S or R)
• LC3: 1.2D + 1.6(Lr or S or R) + (1.0L or 0.5W)
• LC4: 1.2D + 1.0W + 1.0L + 0.5(Lr or S or R)
• LC5: 0.9D + 1.0W
• LC7: 0.9D + 1.0E

Notice something critical in LC5 and LC7: the dead load factor is 0.9, not 1.2 or 1.4. That's because these are uplift and overturning combinations. The dead load is helping you resist the lateral force, so the code lets you use only 90% of it. If the wind or seismic overturning moment exceeds 90% of the stabilizing dead load moment, you need anchorage. Let me run a real example. Our 25-meter silo from the seismic calculation:

Load Combination	Governing Scenario	Max Shell Stress	Foundation Demand
LC2 (Dead + Live + Pressure)	Full grain load	148 MPa	3,200 kN (vertical)
LC4 (Dead + Wind + Live)	70 m/s wind, full load	186 MPa	3,450 kN vert + 1,100 kN horiz
LC7 (0.9D + Seismic)	MCE-level earthquake	212 MPa	2,800 kN vert + 1,880 kN horiz

The seismic combination governs. It always does in high-seismic zones — and the combination with 0.9D is the one that'll try to tip your silo over. Now add thermal stress to the picture. If you're in a region with both high seismicity and large temperature swings (California, Turkey, Chile, parts of China), the thermal hoop stress adds to the seismic hoop stress in the shell. They're both acting in the same direction on certain elements. You need to check the combined stress state, not just each load case independently. For more detail on material behavior under combined loading, see our article on fatigue analysis for cyclic loads on silo shells.

Frequently Asked Questions

What wind speed should I design a silo for?

It depends on your location and applicable code. In the United States, ASCE 7-22 provides risk-based wind speeds from 115-195 mph (ultimate) depending on the building's risk category and location. A grain silo (Risk Category II) in coastal Texas might need 140 mph ultimate wind speed, while the same silo in inland Kansas might only need 115 mph. Always check the local wind speed map and multiply by the appropriate exposure and gust factors. For tropical regions, reference local meteorological data — typhoon or cyclone wind speeds can exceed 170 km/h.

How do I calculate seismic loads for a silo filled with grain?

Per ASCE 7 Chapter 15, treat the stored material using the impulsive-convective model. The impulsive component (roughly 60-80% of the fill mass) moves rigidly with the silo walls. The convective component (the remaining 20-40%) sloshes and has a longer natural period. Calculate the base shear as V = Cs × W, where Cs comes from the building period and site class, and W is the total effective seismic weight (structure plus grain). For most practical silos, the fundamental period is under 0.5 seconds, placing them in the short-period range where seismic coefficients are highest.

Can thermal stress really damage a steel silo?

Absolutely. In climates with 40-50°C daily temperature swings, thermal hoop stress alone can reach 80-120 MPa — which is 35-50% of the yield strength of common structural steel grades. When combined with internal pressure from stored material and external wind or seismic loads, thermal stress can push the total stress state past the design limit. This is especially dangerous for thin-shell silos (3-4mm plates) in the upper courses where external pressure from wind is also highest.

What is the difference between ultimate and service wind speed in silo design?

Ultimate wind speed (used in LRFD design per ASCE 7-16 and later) includes all load factors within the speed itself. Service wind speed (used in older ASD methods) is lower and requires separate load factors. For example, a location with 120 mph service wind speed might have 140 mph ultimate. The structural demand on the silo is the same either way — the difference is just in how the safety factors are applied in the calculation.

How does silo height affect wind load design?

Significantly. The velocity pressure exposure coefficient (Kz) increases with height — at 15 meters in open terrain it's roughly 0.85, but at 25 meters it's approximately 1.07. That's a 26% increase in wind pressure. Additionally, taller silos have lower natural frequencies, making them more susceptible to dynamic effects like vortex shedding. A 30-meter silo in a 60 km/h steady wind may experience oscillating lateral forces that a static analysis completely misses. For silos above 20 meters, dynamic wind analysis is strongly recommended.

Do concrete silos handle seismic loads better than steel silos?

It depends on the seismic zone. Concrete silos have higher mass (which increases seismic demand) but also greater inherent damping and better energy dissipation. In moderate seismic zones (SDS < 0.5g), concrete silos can be more economical because the thicker walls resist buckling without additional stiffening. In high-seismic zones (SDS > 0.8g), steel silos can actually perform better because their lower mass reduces the base shear, and modern seismic detailing (proper anchorage, ductile connections) gives them the toughness they need. The answer is always "it depends" — but run the numbers for your specific site.

When should I add expansion joints to a silo?

For steel silos in climates where the temperature range exceeds 40°C seasonally, expansion joints or slip connections should be considered for silos taller than 15 meters. The total axial expansion for a 25-meter steel silo over a 50°C range is approximately 15mm — enough to induce significant stress at fixed connections if not accommodated. Concrete silos have lower thermal expansion coefficients but are more susceptible to thermal cracking, which requires reinforcement detailing to control crack widths below 0.3mm per most design codes.