Cuttings and Embankments



ALTO HSR · Citizen Research · Infrastructure Risk

Cold-Climate Infrastructure Risk:
Cuttings vs Embankments

Why subgrade form choice is a safety-critical decision for the Ontario–Quebec corridor — and why Canada’s winters make it even more consequential.
⚠ What this means for the consultation

Alto’s current public consultation materials do not specify what proportion of each corridor alignment would require deep cuttings through frost-susceptible terrain. This information is essential for risk-informed public feedback. Communities along the proposed corridors cannot evaluate safety and lifecycle cost implications if fundamental subgrade geometry choices have not been disclosed.

The consultation closes April 24, 2026. This brief summarises peer-reviewed engineering research and applies it directly to the specific geological and climatic conditions of the Alto southern and northern corridor options.

Key Findings
  • Cold-climate HSR research worldwide shows deep cuttings are significantly more dangerous than embankments in snowy, freeze-thaw environments.
  • Frost heave in cut sections is consistently double to triple that of embankment sections on the same line — threatening track geometry tolerances critical for 300 km/h operation.
  • Cuttings act as snowdrift traps, producing an ‘M-shaped’ accumulation directly on the railhead — a derailment risk at high speed.
  • The Alto southern corridor traverses Leda clay and karst terrain where cutting slopes pose catastrophic liquefaction and sinkhole risks not present in embankment construction.
  • Ottawa averages ~90 freeze-thaw cycles per year — far exceeding the climates where most HSR cold-weather research was conducted — making the risk premium for cuttings even greater in Canada.

Definition

What is a railway cutting — and why does it matter?

Railway Cutting — Definition

A cutting is a section of railway excavated below the natural ground level

When a rail line passes through terrain that is higher than the desired track level, engineers excavate downward to create a trench, or “cutting,” through which trains run below the surrounding landscape. The walls on either side may be bare rock, stabilised soil, or reinforced slopes. Cuttings are common in hilly terrain, and proponents often argue they reduce visual impact — but in cold climates, this geometry creates a set of physical hazards that do not apply to the alternative: the embankment.

An embankment is the opposite — a raised platform of compacted fill material on which trains run above the natural ground level. Where an embankment sheds water, snow, and ice away from the track by gravity, a cutting concentrates all of these toward it.

Michael Schabas, author of a January 2026 Senate submission promoting a southern HSR corridor through Eastern Ontario, answered a question posted to his website, noting that “probably about a third will be in cutting.” This brief asks what that means for safety and lifecycle cost in Canada’s winter climate.

Reuters · Photojournalism · January 2026
Two high-speed trains derail and crash in Spain
A Renfe Class 120 trainset in a cutting — one of the trains involved in the January 2026 derailment. 41 people died. The confined geometry of a cutting limits emergency access and concentrates debris on the track.
View full photo coverage at Reuters →

~⅓
Southern Corridor
estimated in cutting (Schabas, 2026)
2–3×
Greater Frost Heave
in cuts vs. embankments — same line
+25%
More Snow Depth
on cutting surfaces vs. embankments
90
Freeze-Thaw Cycles/yr
Ottawa — among highest globally

Peer-Reviewed Evidence

What the engineering literature consistently shows

The following findings draw on studies from operating high-speed rail lines in frozen-ground conditions — principally China’s Harbin-Dalian Passenger Dedicated Line and Scandinavian HSR networks. These are not theoretical projections; they are measurements from real infrastructure operating in winter conditions broadly comparable to Eastern Ontario.

2.1  Snowdrift accumulation: cuttings as traps

Wind-tunnel studies and field monitoring from China’s Xinjiang railways and Scandinavia consistently document a striking asymmetry:

Cuttings produce an ‘M-shaped’ snow accumulation profile. As wind decelerates entering the confined space, particles settle directly on the track surface — on both the windward and leeward shoulders. Snow depth on cutting pavement is typically 25% greater than on equivalent embankment surfaces in the same wind-snow regime.

Embankments produce a ‘U-shaped’ profile, with accumulation at the leeward slope foot — away from the track. The geometry does not concentrate drift onto the railhead.

Key finding (Frontiers in Earth Science, 2022; Journal of Arid Land, 2023): studies specifically comparing cutting and embankment snow regimes conclude that ‘the impact of snowdrift disasters on road cuttings is more serious than that of embankments,’ both in snow depth on the running surface and in visibility reduction.

At 300 km/h, the mechanical energy of impact with a snow obstruction scales with velocity squared. A snowdrift event that would inconvenience a 160 km/h Via Rail corridor becomes a potential catastrophic failure on an HSR line. 1 2

2.2  Frost heave: the cut-section penalty

The most comprehensive real-world data come from the Harbin-Dalian Passenger Dedicated Line in northeast China — the world’s first HSR in seasonally frozen ground, comparable in freeze-thaw intensity to Eastern Ontario. Six years of continuous monitoring at matched cut and embankment sections on the same line produced unambiguous results:

Maximum frost heave in the cut section: 4.52–9.18 mm per season.
Maximum frost heave in the embankment section: 1.86–5.28 mm per season.
Frost heave ratios in the top gravel layer (0–0.5 m): 1.20% in cuts vs. 0.63% in embankments — nearly double.

Miao et al. (2020): ‘frost heave in the cut section was much more serious than that in the embankment section, although the frozen depth of the road cut was approximately 20 cm shallower.’

For HSR at 300 km/h, track geometry tolerances are extremely tight: typically ±2 mm vertical deviation over a 10-metre chord for slab track. A frost heave event reaching 9 mm in a cut section creates a track irregularity that generates dangerous dynamic wheel-rail forces. This is not a maintenance inconvenience — it is a safety threshold breach. 3 4

Critical Finding

Frost heave in cut sections on the Harbin-Dalian HSR was consistently 2–3× greater than in embankment sections on the same line. Eastern Ontario’s freeze-thaw intensity (~90 cycles/year in Ottawa, vs. ~60–70 in Harbin) makes this penalty even more severe for Alto.

2.3  Slope stability and Leda clay: a Canadian-specific catastrophic risk

Every cutting creates two slopes. In stable crystalline rock, these slopes remain predictable over a 100-year asset life. The risk profile in the southern corridor’s glacial sediment terrain is fundamentally different.

Leda clay (Champlain Sea clay / quick clay) underlies the valleys between Smiths Falls and Ottawa through which the Alto southern corridor would pass. Its properties are uniquely dangerous: when disturbed by excavation, vibration, groundwater pressure change, or pore-water pressure increase, it can liquefy catastrophically — transitioning from a particulate solid to a flowing fluid.

Trigger mechanisms include: excavation (inherent to cutting construction), cyclic train-induced vibration (inherent to HSR operation), rapid snowmelt increasing pore-water pressure, and freeze-thaw cycling altering soil fabric.

Over 250 documented Canadian landslides are attributed to Leda clay, including the Lemieux, Ontario landslide (1993, 17 hectares consumed) and the Saint-Jean-Vianney, Quebec disaster (1971, 31 killed, entire town relocated). The 2016 Rideau Street sinkhole in Ottawa was partly attributed to Leda clay destabilised by nearby tunnel construction vibration — exactly the loading regime HSR would impose on a permanent, perpetual basis.

Natural Resources Canada researcher Didier Perret: “When not disturbed, these clays behave very well, but when they are disturbed they behave like fluid.” Deposits can be 15 to 100 metres deep and are typically located below a thin topsoil layer, making visual identification impossible.

A cutting through Leda clay terrain creates a permanent, exposed slope in material that is intrinsically unstable under dynamic loading. The vibration energy of a 300 km/h train propagates into surrounding soils. Over decades of operation, this constitutes a sustained destabilisation regime. Embankments over the same terrain do not create slopes in the sensitive clay; they load it as a foundation — manageable through appropriate pile or raft design. 7 9

2.5  Snow removal and maintenance

Mechanical snow removal from a cutting requires specialised equipment that can operate in a confined space — and removed snow has nowhere to go, requiring loading and transport. Embankment clearance is typically a single-pass operation with swept material descending the slope.

On a 1,000-km network experiencing simultaneous snowfall events — not uncommon in Eastern Ontario — a delay cascade triggered by cutting snow accumulation at one point can propagate through the entire timetable. Scandinavian HSR winter performance studies (Kloow, 2011) document that operational problems increase non-linearly with the duration and geographic extent of winter events. 6

2.6  Emergency egress and rescue

High-speed rail carries hundreds of passengers per train at speeds where incidents, if they occur, are likely to be serious. A train stopped in a deep cutting with 4-metre retaining walls presents a fundamentally different rescue scenario than a train stopped on an embankment: passengers cannot easily self-evacuate laterally, rescue vehicles cannot readily access the train, and in a winter context with snowdrift, the cutting becomes a trap.

This is not a hypothetical risk — winter operational stoppages in cuttings are documented in Scandinavian experience as among the most operationally difficult incidents to manage. 6

Alto-Specific Context

Why these risks are amplified for the southern corridor

The peer-reviewed research summarised above was conducted in China and Scandinavia. The Ontario-Quebec corridor’s conditions are in key respects more demanding than those research environments — and the southern corridor combines multiple compounding geological risk factors largely absent from the northern corridor.

Southern Corridor — Compounding Risks

  • Leda clay along Ottawa–Smiths Falls segment: cutting slopes prone to catastrophic liquefaction under train vibration
  • Karst limestone in the Napanee Plain: cutting through limestone caprock concentrates drainage into dissolution channels, accelerating sinkhole risk
  • Glacial till of variable composition: unpredictable frost-heave behaviour across short track distances — the differential heave most dangerous to 300 km/h operation
  • Lake Ontario lake-effect snow: increasingly severe as lake ice coverage declines, producing sudden, localised, extremely heavy snowfall — the specific trigger for the most dangerous cutting snowdrift scenarios

Northern Corridor — Comparative Advantages

  • Precambrian crystalline rock: rock cuttings do not create Leda clay liquefaction hazards; slopes stable once properly benched and sealed
  • Negligible frost heave: low-permeability crystalline rock suppresses the pore-water migration mechanism that drives frost heave in clays and tills
  • No karst dissolution channels: drainage predictable and surface-dominated; sinkhole risk absent
  • Ideal fill material: rock excavated from Shield cuttings produces well-graded, non-frost-susceptible crushed rock — greatly reducing frost heave risk in embankment sections
  • Stable slope geometry: once engineered, rock slopes in competent Precambrian terrain remain predictable over a 100-year asset life

Canada is not China or Scandinavia

The Harbin-Dalian research was conducted at ~55–65 freeze-thaw cycles per year. Stockholm experiences ~45–55. Ottawa averages ~90. Each cycle imposes stress on subgrade, slopes, and rail infrastructure. Eastern Ontario’s shoulder-season wet snow — combined with rapid temperature drops common in the Frontenac Arch region — creates ideal conditions for ice-laden snow accumulation in cuttings, the hardest form to clear mechanically. The risk premium for cuttings documented in the international literature is a floor, not a ceiling, for the Ontario–Quebec corridor. 3 5

Risk Comparison Matrix

Differential risk profile: cuttings vs embankments in cold-climate HSR

All ratings reflect conditions along the Ontario–Quebec corridor — not general temperate-climate assumptions. Ratings are based on peer-reviewed monitoring data from operating cold-climate HSR lines.

Risk Category Deep Cutting Embankment
Snowdrift Accumulation CRITICAL — ‘M-shaped’ trapping geometry concentrates snow directly on railhead Manageable — ‘U-shaped’ profile; snow fences effective; drift away from track
Frost Heave (subgrade) HIGHER — soil moisture retention; 2–3× greater heave than embankments on same line LOWER — gravity drainage; fill material drier; heave within tolerances
Slope / Mass Failure CRITICAL (southern corridor) — Leda clay liquefaction under cyclic train vibration LOWER — fill slope load manageable via pile/raft design; no exposed quick-clay
Drainage & Meltwater POOR — pooling toward track; ice formation on switches and ballast GOOD — gravity-assisted runoff away from track; lower switch failure risk
Snow Clearing Access DIFFICULT — confined space; removed snow must be loaded and transported STRAIGHTFORWARD — single-pass sweep; material descends slope
Freeze-Thaw Cycling ELEVATED — deeper soil moisture; exposed slopes on all sides MODERATE — drier fill material; reduced water infiltration pathways
OCS Ice Accretion SEVERE — sheltered microclimate promotes rime ice; colder, stiller air MODERATE — wind dispersal reduces ice accretion on catenary
Emergency Egress CONSTRAINED — retaining walls block lateral exit; rescue vehicle access limited ACCESSIBLE — open sides; lateral evacuation possible; vehicles can approach
Karst / Sinkhole Risk AMPLIFIED (southern corridor) — concentrated drainage accelerates dissolution; caprock breached LOWER — distributed drainage; limestone cap preserved; reduced dissolution

Ratings based on peer-reviewed monitoring data from operating cold-climate HSR lines, applied to Ontario–Quebec corridor conditions. 1 2 3 4 5 6 7

Design and Consultation Implications

Three things Alto must address before the consultation closes

Subgrade form must be treated as a risk-scored variable. Alto’s current public consultation materials do not provide corridor-by-corridor analysis of the proportion of each alignment requiring deep cuttings through frost-susceptible terrain. This information is essential for risk-informed public feedback.
Cold-climate design standards must be explicitly specified. The Harbin-Dalian experience demonstrates that ‘standard’ HSR design assumptions derived from temperate European practice are insufficient for Canadian winters. The NRC Canada freeze-thaw cycle data for the Ontario-Quebec corridor must be a design input, not an afterthought.
Route selection should include subgrade form as a scoring criterion. If the northern corridor can achieve a significantly higher proportion of embankment construction on stable, non-frost-susceptible rock fill compared to the southern corridor’s requirement for cuttings through Leda clay and glacial till, this differential should appear explicitly in Alto’s route scoring methodology.

“A deep cutting through Leda clay, traversed by 300 km/h trains, exposed to 90 freeze-thaw cycles per year, subject to lake-effect snowdrift, and draining into karst watersheds is not a manageable engineering challenge. It is a liability cascade.”

Submit your comments by April 24, 2026 →

Key References

Sources and further reading

1
Li, P.X. et al. (2022). “Analysis of the influence of snow fences on snow redistribution under snow-drifting in railway cuttings.” Cold Regions Science and Technology, 196, 103520.
2
Yang, Y. et al. (2022). “Study on the evolution process of snow cover in wind-induced railway embankments and the control effect of snow fences.” Applied Sciences, 15(11), 6057.
3
Miao, Q. et al. (2020). “Comparing frost heave characteristics in cut and embankment sections along a high-speed railway in seasonally frozen ground of Northeast China.” Cold Regions Science and Technology, 170, 102921.
4
Zhang, S. et al. (2016). “Analysis of frost heave mechanisms in a high-speed railway embankment.” Canadian Geotechnical Journal, 53(3), 520–529.
5
Liu, H. et al. (2016). “Effect of structures and sunny-shady slopes on thermal characteristics of subgrade along the Harbin-Dalian Passenger Dedicated Line.” Cold Regions Science and Technology, 123, 14–21.
6
Kloow, L. (2011). High-Speed Train Operation in Winter Climate. Royal Institute of Technology, Stockholm.
7
Locat, A. et al. (2017). “Sensitive clay landslides in Canada.” Landslides in Sensitive Clays. Springer, Dordrecht.
8
Stipanovic, I. et al. (2013). “A risk assessment methodology for weather-related failures of railway infrastructure.” International Journal of Railway, 6(3), 99–107.
9
Perret, D. (Natural Resources Canada). Leda clay sensitivity and eastern Canada landslide inventory.
10
Palin, E.J. et al. (2021). “Implications of climate change for railway infrastructure.” WIREs Climate Change, 12(5), e728.
11
Schabas, M. (2026). “Alto High-Speed Rail Conceptual Design and Business Plan.” Senate Subcommittee on Transport and Communications, January 2026.
12
Reuters (2026). High-speed train derailment, Spain. Reuters coverage