Control piles back in service: how UNAM engineers restored the foundations of Mexico City’s Metropolitan Cathedral

Stratigraphic profile beneath the front of the Cathedral and the Sagrario, with cone resistance profiles SCE-2, SCE-1 and SCE-6 showing fill, surface crust, upper clay (FAS), the hard layer at about 38–40 m, lower clay (FAI) and deep deposits. Source: Botero-Jaramillo et al. (2026), TC301 Athens.

After years without maintenance and the 2017 earthquakes, a 2020–2021 geotechnical program at Mexico City’s Metropolitan Cathedral re-read the piezometers, repeated CPTs, load-tested and reconditioned control piles, and checked them against Tamez, Zeevaert and NTC 2017 criteria.

Key facts

Cathedral: 126.7 m × 60.4 m, total weight about 1,270 MN, average bearing pressure about 166 kPa
Of 417 control piles, only 28% carry load – 3,912 t (about 38.4 MN) in total, just 3% of the structure’s weight
Six piles load-tested in 10 t increments up to 120 t; operating loads re-set at 10, 30, 40 and 90 t after maintenance
Pore pressures hydrostatic from the 7.5 m water table to the hard layer at 38 m, with drawdown between 38 and 49 m unchanged over 21 years
Zeevaert safety factors of 1.97–2.46 (0.40 m piles) and 2.33–2.84 (0.45 m piles) for the test-program loads

Mexico City’s Metropolitan Cathedral is one of the best-known cases of a heritage structure fighting differential settlement on soft lacustrine clay. Between 1989 and 2001 a major geotechnical rehabilitation sharply reduced both the differential settlements and their rate of growth, and structural and foundation reinforcement continued into the following decade. Funding then dried up, maintenance and automated monitoring stopped, and the strong earthquakes of 2017 caused damage classified as low severity. In a paper for the 4th International Symposium on Geotechnical Engineering for the Preservation of Monuments and Historic Sites (ISSMGE TC301) in Athens, Eduardo Botero-Jaramillo, Efraín Ovando Shelley, Alexandra Ossa López, María C. Madrigal Madrigal, Renata A. González Rodríguez and A. Roberto Sánchez Ramírez of the Institute of Engineering of the National University of Mexico (II-UNAM) describe the geotechnical work carried out in 2020–2021 to repair that damage and bring the foundation system back into service.

Figure 1. Stratigraphic profile beneath the front of the Cathedral and the Sagrario, with cone resistance profiles SCE-2, SCE-1 and SCE-6 showing fill, surface crust, upper clay (FAS), the hard layer at about 38–40 m, lower clay (FAI) and deep deposits. Source: Botero-Jaramillo et al. (2026), TC301 Athens.

A heavy monument on a pre-Hispanic island

The Cathedral, built between 1571 and 1813, has five naves, a 65 m high central dome on four columns and two 60 m bell towers. It is 126.7 m long and 60.4 m wide, weighs about 1,270 MN and transmits an average bearing pressure of about 166 kPa. The adjoining Sagrario Metropolitano (1749–1768), a Greek-cross building 47.7 m square weighing about 300,000 kN, imposes about 132 kPa.

The complex stands on part of the natural island occupied by the Mexica, and its foundations contain remains of pre-Hispanic structures; the Sagrario overlies remains associated with the pyramid of Tonatiuh. The builders drove about 22,500 short timber piles and placed a masonry platform 140 m by 70 m, averaging 0.90 m thick but deliberately thickened towards the south to compensate for settlements already observed during construction. On this platform sits a grid of masonry tie-beams about 3.5 m high and 2.5 m wide, up to 127 m long, which carries the walls, pilasters and columns.

Figure 2. The original foundation system: plan of the 140 m × 70 m rock-fill platform with its grid of 2.5 m × 3.5 m inverted beams (a), section showing crypts, rock fill, timber piles and buried pyramid remains (b), and the Sagrario with its short wood piles (c). Source: Botero-Jaramillo et al. (2026), TC301 Athens.

The subsoil consists mainly of highly compressible lacustrine soft clays formed from alluvial material and volcanic ash, interbedded with ash, pumice and weathered volcanic, fluvial, alluvial and glacial deposits. The result is a sequence of soft clay strata separated by stiff lenses of clayey–sandy silt, desiccation crusts and basaltic or pumiceous sands. Regional subsidence caused by more than a century of groundwater extraction has made the damage worse.

Piezometry and CPTs two decades on

In July 2021 the team carried out a CPTu with pore-pressure dissipation tests and read the piezometer stations and an observation standpipe. Compared with 1999 records, the depth to the water table has not changed significantly in 21 years. From the present groundwater level at about 7.5 m down to the hard layer at 38 m, pore pressures remain essentially hydrostatic. Between 38 m and 49 m they show drawdown relative to hydrostatic, but this too has not changed appreciably over two decades – consistent with earlier hydrogeological studies placing the main drawdown driving subsidence in the Historic Centre at depths greater than 50 m.

A new CPT (CPT-2_Ing, 2021) was compared with soundings made at the base of the West Tower in 1989 and 2001. The 2021 profile shows higher cone resistance, which the authors attribute to soil stiffening caused by subsidence associated with groundwater withdrawal. Water levels in the shafts showed minimal fluctuations overall, although differences of up to 2 m in shafts L-09, L-12 and L-16 call for a site review to rule out leakage in the piezometer tubing.

Figure 3. Left: pore pressure with depth from piezometers and CPTu dissipation tests in July 2021 compared with 1999 data, essentially hydrostatic to the hard layer. Right: cone resistance from CPT-2_Ing (2021) against the 1989 and 2001 soundings, with higher values in 2021. Source: Botero-Jaramillo et al. (2026), TC301 Athens.

Load testing and reconditioning the control piles

The Cathedral has more piles than any other building in the complex, but of its 417 control piles – piles fitted with load frames and deformation cells that apply and hold a controlled axial load to limit differential settlement – only 28% were carrying load. Together they supported 3,912 t (about 38.4 MN), just 3% of the structure’s total load, and 80% carried no more than 50 t (about 490 kN).

Years without maintenance had taken their toll, especially in the south-west sector: bolts, plates and reaction frames were badly corroded, and some of the hardwood (caobilla) blocks that keep a constant load on the pile heads were deformed beyond allowable limits or rotten. The works included maintenance of the reaction frames, trimming of pile heads with significant apparent protrusion, replacement of the timber blocks, application of calculated loads and load testing of selected piles.

Figure 4. Restored reaction frames on control piles 15, 16 and 17 beneath the Cathedral, part of the 2020–2021 foundation maintenance. Source: Botero-Jaramillo et al. (2026), TC301 Athens.

II-UNAM designed and supervised load tests on six piles along the south façade, each initially carrying 25 t (about 245 kN). They were loaded in 10 t steps up to 120 t (about 1.18 MN), with displacements recorded at each step, then unloaded and set to their design operating loads. The tests showed how load was carried along the shaft and at the toe, and so whether each pile was socketed into the hard layer or fully embedded in clay. A loading protocol for the southern zone, intended to reduce the settlement rate there, took account of available pile capacity, recorded pile emersions and the properties of the timber blocks. Where installed blocks turned out to have poorer properties than specified, the problem was remedied, then and in subsequent maintenance. Final loads of 10, 30, 40 and 90 t (about 98 to 883 kN) were set from calculations of column and façade out-of-plumbness.

Checking the piles: Tamez, Zeevaert and NTC 2017

The team checked end-bearing and friction control piles of 0.40 m and 0.45 m diameter in the southern sector against the ultimate limit state of the Mexico City Complementary Technical Standards for foundations (NTC, 2017), together with the criteria of Tamez (2001) and Zeevaert (1983). Under Tamez’s criteria, the slab must settle at the same rate as the surrounding ground, and the tip capacity must exceed the deformation-cell yield load plus shaft capacity times a safety factor, so that the pile tip does not punch into the resistant stratum. Zeevaert’s approach adds the negative skin friction from surface loading and subsidence to the load from the control device. Because the properties of the hard layer are uncertain, friction angles of 30°, 40° and 45° were considered, with at least 40° expected.

On that basis, piles of both diameters bearing on the hard layer have enough tip capacity under Tamez’s criterion. Zeevaert safety factors range from 1.97 to 2.46 for the 0.40 m piles and from 2.33 to 2.84 for the 0.45 m piles. The NTC check depends on how far the piles are embedded in the frictional stratum. With zero embedment, 0.40 m piles fail the check at 50, 75 and 95 t and 0.45 m piles at 75 and 95 t. With minimum embedments of 0.70 m and 0.45 m respectively, both satisfy the standard under all proposed loads. Friction piles satisfy NTC 2017 only for loads up to 25 t, provided a minimum length is achieved.

Why it matters for practice

The authors stress that the control piles do not correct or govern the behaviour of the Cathedral and the Sagrario; they contribute to the stability of the foundation and structure. The case shows how quickly a carefully engineered settlement-control system deteriorates without maintenance, and why pile embedment and hard-layer strength matter so much in any capacity check. It also shows the value of long piezometric and CPT records for separating regional subsidence effects from local behaviour. The team calls the continued monitoring of piles and instrumentation, periodic levelling and a dedicated annual maintenance budget “essential and urgent”, and argues that a broader programme is needed to keep the complex in good condition.