Bonded vs Unbonded Post-Tensioning: Which System for Your Project?
Ask a non-specialist engineer in West Africa which post-tensioning system to specify and you will hear one of two wrong answers: "whichever is cheaper" or "whichever the contractor offers." Both miss the point. Bonded and unbonded post-tensioning are structurally distinct systems with different ultimate flexural capacity, ductility, fire response, corrosion protection mechanisms and construction logistics. Default to unbonded mono-strand on a transfer slab and you get a structure that satisfies serviceability but underperforms at ULS. Insist on bonded multi-strand in a 7 m residential one-way slab and you pay for grouting plant and inspection hold points for no engineering benefit. The choice matters, and it is made at concept design, not at tender.
This article is written for structural engineers, design firms and consultants specifying PT concrete in West Africa. It compares bonded vs unbonded post-tensioning on five engineering axes — flexural behaviour, ductility, durability, fire and constructability — sets out an applications matrix, and shares cost ballparks measured across BEPCO's 2026 projects in Côte d'Ivoire, Ghana, Nigeria and Senegal. Where Eurocode 2, ACI 318 and PTI guidance differ, we cite the source.
By BEPCO engineers, specialists in post-tensioned concrete across 11 West African countries for 15+ years. Last updated: May 2026.
What is bonded post-tensioning?
Bonded post-tensioning houses multiple high-strength prestressing strands inside a corrugated metal or HDPE duct cast into the concrete. After release strength is reached, the strands are stressed against anchorages at each end. Once stressing and lift-off are recorded, the duct is filled with cementitious grout under controlled pressure. When the grout hydrates, it forms a strain-compatible mechanical bond between every strand and the concrete along the full tendon length.
Multi-strand tendons and duct sizes
A typical bonded tendon in a building or short-span bridge contains 4 to 19 strands of 13 mm or 15.7 mm Y1860 strand, in a flat or oval duct for slabs and a circular duct for beams. Larger infrastructure tendons can carry up to 55 strands in a circular duct of 130 mm or more. A 19-strand stressing anchorage transfers about 4,000 kN to the concrete and demands local zone reinforcement designed to PTI M50.3 or fib bulletin 33 rules.
Why grouting changes the structural behaviour
The grout does two things at once. First, it provides the corrosion protection envelope: the alkaline environment passivates the steel and the duct separates the tendon from chlorides diffusing through cover. Second, it bonds strand to concrete so strain in the strand at any section equals strain in the surrounding concrete (modular ratio adjusted). That strain compatibility gives bonded post-tensioning its distinctive ULS behaviour — strand stress at the critical section can rise close to ultimate tensile strength, and a local strand failure does not unload the entire tendon.
What is unbonded post-tensioning?
Unbonded post-tensioning uses individual mono-strands, factory-encapsulated for life. A standard unbonded tendon is a single 13 mm or 15.7 mm Y1860 strand, coated with corrosion-inhibiting grease, then sheathed in an extruded HDPE jacket of 1.0-1.5 mm wall thickness. It is delivered on coils, cut, threaded through anchorages, profiled on chairs and cast in. After release strength, each mono-strand is stressed individually with a 15-30 kg light jack. There is no grouting step.
The strand floats inside its sheath
Because grease prevents adhesion between strand and sheath, the strand is mechanically free to slide. Strand strain — and tendon force — is essentially uniform between anchorages, modified only by friction losses (typical wobble 0.0033 per metre, curvature 0.06-0.07 per radian for greased mono-strands). At ULS, the strand stress increase above effective prestress is small and is calculated in ACI 318-19 section 20.3.2.4 as a span-to-depth-dependent term typically capped at 210 MPa above effective prestress.
Where the corrosion protection comes from
With no grout, the corrosion protection envelope is grease film, HDPE sheath and concrete cover to the sheath. PTI M10-series specifications define encapsulated systems where the anchorage is fully sealed and the strand is electrically isolated end-to-end — equivalent to fib Protection Level 2 or 3. For coastal and aggressive exposure, BEPCO specifies fully encapsulated PTI Class C systems by default; the marginal cost is small and the durability margin large.
Engineering differences: where the two systems diverge
The behaviour gap is widest at ULS and narrowest in service. Both systems compress concrete along identical force paths, control deflections similarly, and operate crack control by precompressing the tension fibre. The differences emerge past first cracking and toward flexural failure.
Stress redistribution after cracking
In a bonded section, when the tension fibre cracks, local strand strain rises sharply because compatibility forces the strand to elongate with the cracked concrete. Strand stress at the crack climbs toward ultimate while strands a metre away stay near effective prestress; load continues to be carried because the high-stress zone is local. In an unbonded section, the same crack causes a small uniform stress increase along the entire tendon — there is no local concentration because the strand slides. The unbonded section reaches its ultimate moment at a lower strand stress than an identical bonded section.
Ultimate flexural capacity and bonded mild steel
Eurocode 2 and ACI 318 both account for this: an unbonded tendon contributes less to ULS capacity than a bonded one of equivalent area. ACI 318-19 requires minimum bonded reinforcement (typically 0.004 Act in positive moment regions of unbonded slabs) precisely to provide ductility and crack control the unbonded tendon alone cannot guarantee — more onerous still for one-way slabs and beams. In bonded systems, the strands themselves provide the ductility, so the mild-steel requirement is reduced or, in some bridge applications, eliminated.
Ductility and progressive collapse
If a single strand fails locally — by section loss from corrosion, or wire fracture under seismic loading — consequences differ. A bonded tendon retains most of its force because bond re-anchors the strand within a transfer length of the failure (typically 1-1.5 m). An unbonded mono-strand that fails or loses its anchorage releases its entire force across the full tendon length. Multi-strand bonded systems therefore have built-in redundancy in the tendon itself; unbonded systems achieve redundancy only through multiple independent strands. This is one reason bonded is mandated in long-span bridges.
Construction differences: what the site team actually does
The site reality is very different — and this is where many cost and programme assumptions break. Unbonded mono-strand is fast, light and forgiving. Bonded multi-strand is heavier, slower per metre and demands a specialist grouting operation with no equivalent in the unbonded workflow.
Installation rates per crew
A four-person BEPCO unbonded crew on a typical office or residential floor places 2,500-3,500 m of mono-strand per day. The same four-person crew on 4-strand or 7-strand bonded tendons in flat ducts places 800-1,400 m per day — roughly a third of the rate, because ducts are heavier, strands must be threaded after duct placement, and connections require taping and inspection at every coupler. Per kN of prestress installed the gap narrows; per linear metre of tendon the unbonded system is much faster on slab geometry.
Stressing sequence and equipment
Unbonded mono-strands are stressed individually with a 25-30 t light jack one operator can carry. Stressing is fast — a 600 m² slab with 80 mono-strands is fully stressed and recorded in half a shift. Bonded multi-strand tendons are stressed all strands at once with a 200-1,000 t multi-strand jack, requiring lifting plant, hydraulic power packs and a more involved lift-off and elongation reconciliation. Crew skill is higher; jack capital cost is significantly greater. BEPCO's bonded multi-strand jacking inventory is one of the largest in West Africa — see our bridges and infrastructure page for representative scope.
Grouting QA: the bonded-system risk that does not exist in unbonded
Grouting is the single most failure-prone step in bonded PT. FHWA guidance updated after 2000 traces most long-term durability failures to grout voids. Best practice — BEPCO standard on every bonded contract — uses thixotropic pre-bagged grout, vacuum-assisted injection where geometry permits, vent management at every high point, and a written grout record per tendon. On critical bridge work we run mock-up tendons before the first production grout. None of this applies to unbonded; the corresponding QA is sheath integrity inspection at receipt and during placement.
Estima el ahorro para tu propio proyecto. Nuestra calculadora gratuita modela el espesor de losa, el ahorro de acero y el impacto en el coste en menos de un minuto. Prueba la calculadora →
Durability differences: how each system survives 50+ years
Both systems can achieve 50-100 year design lives when correctly specified, supplied and installed. Both can also fail prematurely when corners are cut. The failure modes and the QA controls differ.
Bonded: grout-dependent corrosion protection
Bonded tendon protection depends on duct integrity, grout completeness and grout alkalinity. A void at a high point — a vent closed prematurely, a duct crushed by misplaced rebar — leaves the strand exposed to moisture and chlorides arriving via later cracking or anchorage weaknesses. Encapsulated HDPE-duct systems with electrically isolated anchorages (PTI PL3) raise the durability ceiling significantly and are the BEPCO default for coastal infrastructure. See our coastal corrosion and durability article for the specific drivers we manage.
Unbonded: sheath-dependent corrosion protection
Unbonded mono-strand protection depends on the HDPE sheath staying intact, grease film staying in place, and the anchorage staying sealed. The first risk is sheath damage during placement — a strand dragged across a sharp form edge or punctured by tie wire. BEPCO crews inspect every metre of sheath at placement and repair damage with PTI-approved sealant before pour. The second risk is anchorage water ingress — addressed by encapsulated anchorages with grease-filled caps. Properly executed, fully encapsulated unbonded systems are PL2/PL3 equivalent with a strong track record in the Florida and Texas coastal markets that share environmental severity with West African coasts.
Fire performance
Bonded tendons perform better in fire for two reasons. First, in a bonded section a localised loss of strand strength affects only the local zone — concrete and bond redistribute load. In an unbonded section, loss of anchorage or sheath integrity at one point releases the entire tendon. Second, the grout adds a small thermal lag relative to the directly-sheathed unbonded strand. For both systems, fire rating is governed by concrete cover — see our fire resistance guide for the cover and rating tables we use.
Cost differences: system cost, labour and equipment
Cost comparisons are misleading when expressed only as supply price per kg of strand. Supply is one input; the larger differences are in installation labour, jacking and grouting plant, anchorages and inspection. The supplied-and-installed ranges below cover 2026 BEPCO projects, with the higher end of each range covering smaller projects, less accessible sites or smaller tendon sizes.
- Unbonded mono-strand, supplied and installed: USD 6.0-10.0 per kg of strand on a typical building floor slab.
- Bonded multi-strand, supplied and installed (4-7 strand tendons): USD 8.0-13.0 per kg, including duct, grout and grouting QA.
- Bonded large multi-strand (12-19 strand bridge tendons): USD 11.0-16.0 per kg, including encapsulated PL3 anchorages.
- Slab system delta (bonded vs unbonded, like-for-like flat slab): bonded adds approximately 10-15 % to the post-tensioning sub-package cost.
For a 5,000 m² flat slab in Lagos, Accra or Abidjan, the 10-15 % delta is USD 7,000-14,000 in extra sub-package cost. On parking or office floors where unbonded is structurally appropriate, the delta is unjustified. On transfer slabs or bridges where bonded is required, it is the price of admission and excellent value for the engineering capability obtained. Our post-tensioning quick-cost calculator takes typology, span and area and returns a system-cost ballpark including the bonded-vs-unbonded choice.
Side-by-side comparison and applications matrix
The two tables below summarise the engineering and construction differences and translate them into the typology-by-typology recommendation BEPCO uses on concept advice.
| Criterion | Bonded multi-strand | Unbonded mono-strand |
|---|---|---|
| Tendon configuration | 4-19 strands in metal/HDPE duct, grouted | Single 13 or 15.7 mm strand, greased, HDPE sheath |
| Stress redistribution after cracking | Strong (local strain compatibility) | Weak (strain uniform along tendon) |
| Ultimate flexural capacity per kN of prestress | Higher | Lower (ACI cap on strand stress at ULS) |
| Bonded mild steel requirement | Reduced | Mandatory (ACI 318 minimum) |
| Ductility and progressive collapse | Better (strand re-anchors via bond) | Requires bonded reinforcement to provide ductility |
| Fire resistance (governed by cover) | Marginally better | Slightly more sensitive to local damage |
| Corrosion protection mechanism | Grout + duct + cover | Sheath + grease + cover (encapsulated PL2/PL3) |
| Installation rate per crew (slab) | 800-1,400 m/day | 2,500-3,500 m/day |
| Stressing equipment | 200-1,000 t multi-strand jack | 25-30 t hand-portable mono-strand jack |
| Grouting required? | Yes - critical QA step | No |
| Repairability | Difficult once grouted | Anchorage replacement possible |
| Inspection | Lift-off limited; tendon condition harder to verify | Lift-off per individual strand |
| Supplied-and-installed cost (USD/kg strand) | 8.0-13.0 (16.0 for bridges) | 6.0-10.0 |
| Specialist contractor capability in West Africa | Limited (BEPCO is one of few full-capability suppliers) | Widely available |
| Structure type | Recommended system | Why |
|---|---|---|
| Residential/office floor slab | Unbonded (80 % of cases) | Spans 7-9 m, ductility from bonded mild steel sufficient, fast crew rates |
| Transfer slab over column transfer | Bonded | High concentrated forces, ULS capacity demand, local stress redistribution critical |
| Deep transfer beam | Bonded | Same as transfer slab; multi-strand tendon geometry suits beam depth |
| Mat foundation / raft | Bonded | Long-life corrosion protection in buried environment, redundancy |
| Bridge superstructure | Bonded (always) | FHWA, Eurocode and PTI guidance: redundancy and ULS demand |
| Multi-storey parking deck | Either - unbonded common | Spans 8-16 m suit either; unbonded faster and cheaper |
| Industrial slab-on-grade | Unbonded mono-strand | No grouting equipment on dispersed sites, easy stressing access |
| Tank or reservoir circumferential | Bonded | Multi-strand circumferential tendons, life-cycle critical |
| Podium over basement parking | Bonded for transfer zones, unbonded elsewhere | Hybrid optimises cost and ULS performance |
From the BEPCO project record
"On a 2025 mixed-use podium in Abidjan, the original tender specified unbonded mono-strand throughout, including the transfer slab over a basement column elimination. Our concept review flagged the transfer zone as inappropriate for unbonded — the calculated ULS moment demand exceeded unbonded capacity by approximately 12 % even with minimum bonded reinforcement at code limits. We proposed a hybrid: 4-strand bonded tendons in the transfer band, mono-strand in the surrounding flat-slab field. The change added 4 % to the PT sub-package and removed a ULS deficiency the original design had not identified. Programme impact was zero — the bonded crew worked one shift parallel to the mono-strand crew on a different bay." — From the BEPCO project record
Frequently asked questions
Can bonded and unbonded tendons be used in the same slab?
Yes, and on transfer-zone projects it is often the most efficient solution. The two systems are independently designed, stressed and inspected; there is no structural interaction provided anchorage zones do not clash. BEPCO has executed several hybrid slabs across Côte d'Ivoire and Senegal in the last three years.
Is bonded post-tensioning available everywhere in West Africa?
Bonded multi-strand capability — heavy multi-strand jacks, grouting plant and trained crews — is concentrated in a small number of specialists. BEPCO is one of the few full-capability suppliers across the 11-country region. The unbonded supplier base is broader, though anchorages and trained crews remain limiting factors in smaller markets.
Does the bonded vs unbonded choice affect the slab depth?
Marginally. In serviceability-governed slabs, both systems deliver similar span-to-depth ratios (typically L/40 to L/45 for two-way flat slabs). In ULS-governed transfer applications, bonded permits a thinner section because more strand stress can be mobilised at ultimate.
What inspection regime applies to each system in service?
Unbonded systems permit lift-off of each strand at any time in service — useful for periodic inspection and forensic investigation. Bonded systems do not permit lift-off after grouting; in-service inspection relies on visual assessment of anchorage zones, GPR or impact-echo for void detection, and instrumented monitoring on critical infrastructure. Our audit and expertise service covers both inspection regimes.
Which system is required by Eurocode 2 for a given application?
Eurocode 2 does not mandate one system per typology — the choice remains the designer's. Eurocode 2 does set different rules: section 5.10 covers prestress losses for both, and the French National Annex imposes additional requirements on grouting and encapsulation. We design West African projects to Eurocode 2 with PTI and fib supplementary guidance, and cross-check against ACI 318 where the client's frame is US-based.
Conclusion: the choice is an engineering decision, not a procurement one
The bonded vs unbonded decision is a clean example of engineering judgement made at concept design. Default unbonded for residential and office slabs, parking decks and most light-industrial. Specify bonded for transfer slabs, deep beams, mat foundations, bridges, tanks and any structure where ULS capacity, redundancy or grouted long-life corrosion protection dominate. Hybrid systems are viable where one zone needs bonded behaviour and the surrounding field does not. The 10-15 % cost penalty on a like-for-like slab is small compared to the consequence of specifying the wrong system.
BEPCO supplies, designs and installs both systems across 11 West African countries, with one of the deepest bonded multi-strand capability bases in the region. If you are at concept design on a project that may benefit from PT — or if a tender specifies a system that seems wrong for the application — contact our engineering team for a one-page review. We work alongside design firms and consultants without competing for design fees; early-stage advice is free for serious enquiries.
Sources and references
- Post-Tensioning Institute (PTI), specifications M50.3 and M10 series — post-tensioning.org
- American Concrete Institute, ACI 318-19 (Building Code) and ACI 423.10R (Bonded post-tensioning report) — concrete.org
- Eurocode 2 (EN 1992-1-1) for prestressed concrete design — eurocodes.jrc.ec.europa.eu
- FHWA bridge post-tensioning durability guidance — fhwa.dot.gov
- fib bulletin 33 (Durability of post-tensioning tendons) and fib bulletin 43 — fib-international.org
Related reading
- Post-tensioning in Nigeria: a developer's guide for Lagos projects
- Construction costs in Ghana 2026: where post-tensioning fits
- Coastal West Africa: corrosion and durability in post-tensioned concrete
- Post-tensioning fire resistance: engineers' and developers' guide
- BEPCO services: post-tensioned slabs and floors
- BEPCO services: beams and long-span structures