Concrete Mix Design – TMS, Margin & Standard Deviation Tool – Updated 2026 BS EN 206 / BS 8500
Calculate the required Target Mean Strength (fcm,target) for concrete mix design from the specified characteristic strength, production standard deviation, and margin. Based on BS EN 206:2013+A2:2021 and BS 8500-1:2015+A2:2019.
The Target Mean Strength (TMS) — also called the required mean strength or aim strength — is the mean compressive strength that a concrete mix must be designed to achieve in order to ensure that the statistical proportion of results falling below the specified characteristic strength (fck) does not exceed the permitted defect rate (typically 5% under BS EN 206). It is the central output of the concrete mix design process and directly determines the required water/cement ratio, cement content, and other mix proportions.
The TMS is always greater than the characteristic strength because any production process has variability — even the best-controlled ready-mix plant will produce some batches slightly weaker or stronger than the mean. By designing the mix to a mean strength above the characteristic strength (by the margin M = kσ), the statistical probability of any individual result falling below fck is limited to the acceptable defect rate.
TMS and mix design resources: BSI – BS EN 206 / BS 8500 | The Concrete Society | The Concrete Centre | British Ready Mixed Concrete Association
| fck,cyl (N/mm²) | BS 8500 Grade | TMS σ=3.0 | TMS σ=4.0 | TMS σ=5.0 | TMS σ=6.0 | TMS σ=7.0 | TMS σ=8.0 |
|---|---|---|---|---|---|---|---|
| 12 | C12/15 | 16.9 | 18.6 | 20.2 | 21.8 | 23.5 | 25.1 |
| 16 | C16/20 | 20.9 | 22.6 | 24.2 | 25.8 | 27.5 | 29.1 |
| 20 | C20/25 (GEN3) | 24.9 | 26.6 | 28.2 | 29.8 | 31.5 | 33.1 |
| 25 | C25/30 | 29.9 | 31.6 | 33.2 | 34.8 | 36.5 | 38.1 |
| 28 | C28/35 | 32.9 | 34.6 | 36.2 | 37.8 | 39.5 | 41.1 |
| 30 | C30/37 (RC30) | 34.9 | 36.6 | 38.2 | 39.8 | 41.5 | 43.1 |
| 32 | C32/40 | 36.9 | 38.6 | 40.2 | 41.8 | 43.5 | 45.1 |
| 35 | C35/43 (RC35) | 39.9 | 41.6 | 43.2 | 44.8 | 46.5 | 48.1 |
| 40 | C40/50 (RC40) | 44.9 | 46.6 | 48.2 | 49.8 | 51.5 | 53.1 |
| 45 | C45/55 | 49.9 | 51.6 | 53.2 | 54.8 | 56.5 | 58.1 |
| 50 | C50/60 | 54.9 | 56.6 | 58.2 | 59.8 | 61.5 | 63.1 |
| Number of Results n | Mean Must Exceed | k1 Value | Notes |
|---|---|---|---|
| 3 | fck + 4.0 N/mm² | — | Fixed margin for initial production (3–14 results) |
| 10–14 | fck + 4.0 N/mm² | — | Fixed margin; use minimum margin until ≥15 results available |
| 15–24 | fck + k1σ | 1.48 | k1σ calculated from rolling standard deviation of group |
| ≥ 25 | fck + k1σ | 1.48 | Full statistical assessment; σ from last 35 results preferred |
Criterion 2 applies independently of sample size: every individual result must be ≥ fck − 4 N/mm². This catches any single excessively low result regardless of the mean. For C20/25 (fck = 20 N/mm²): every cube or cylinder result must be ≥ 16 N/mm². For C30/37: every result must be ≥ 26 N/mm².
| Strength Class | Initial Margin (σ unknown) | Corresponding TMS (C30/37 example) | Notes |
|---|---|---|---|
| fck < C35 | +4.0 N/mm² (fixed) | 34.0 N/mm² | Applied when <15 test results available; conservative |
| fck ≥ C35 | +6.0 N/mm² (fixed) | 41.0 N/mm² | Higher fixed margin for stronger concretes with greater variability |
| Any class (k=1.64, σ=4) | +6.6 N/mm² (statistical) | 36.6 N/mm² | Standard statistical margin for well-controlled plant |
| Any class (k=1.64, σ=6) | +9.8 N/mm² (statistical) | 39.8 N/mm² | Less controlled production; higher TMS required |
| Target Mean Strength fcm (N/mm²) | Approx. w/c Ratio | Approx. Cement Content (kg/m³) | Notes |
|---|---|---|---|
| 20–22 | 0.65–0.70 | 260–290 | GEN3/C20 domestic mix; not suitable for XC3+ exposure |
| 25–27 | 0.58–0.65 | 280–310 | C25/30 typical mix |
| 30–32 | 0.52–0.58 | 300–340 | C28/35 to C30/37; XF1, XC3 compliance range |
| 35–37 | 0.46–0.52 | 320–360 | C30/37 to C32/40; XC4, XS1 compliance range |
| 40–42 | 0.42–0.48 | 340–380 | C35/43 to C40/50; XF4, XS2/3 range |
| 45–50 | 0.38–0.44 | 360–420 | High strength; structural applications |
| 50–60 | 0.34–0.40 | 380–450 | C50/60+; may require silica fume or high-range WRA |
w/c ratios above are approximate guidelines for CEM I Portland cement at standard workability (S3 slump) at 20°C curing. SCMs (GGBS, PFA) alter the relationship; hot weather reduces strength for a given w/c; retarders may be needed for long transport. Always use proper mix design procedures (BRE/DoE method or equivalent) and plant-specific calibration data rather than this table for production mix design.
The characteristic strength (fck) is a minimum guarantee — only 5% of results should fall below it. The target mean strength (TMS) is what the mix must actually be designed to achieve on average. TMS is always higher than fck by the margin M. For C30/37 at a well-controlled plant (σ=4): TMS ≈ 36.6 N/mm², not 30 N/mm².
BS EN 206 specifies that the characteristic strength fck is the value below which no more than 5% of results from a defined population are expected to fall. For a normal distribution, this corresponds to the mean minus 1.645 standard deviations (k ≈ 1.64). Designing to TMS = fck + 1.64σ achieves this statistical guarantee.
Standard deviation σ is the most important production-control parameter in TMS calculations. A plant reducing σ from 6 to 4 N/mm² can reduce TMS by 3.3 N/mm² (= 1.64×2) — allowing a lower w/c ratio to be used, reducing cement content by approximately 20–30 kg/m³, and saving approximately £5–£10/m³ of concrete cost. Process improvement and consistent quality control directly save money on every cubic metre produced.
In EC2/BS EN 206 notation, characteristic strength is specified as cylinder strength (fck,cyl), with the cube equivalent given after the slash: C30/37 has cylinder strength 30 N/mm² and cube strength 37 N/mm². The relationship: fck,cube ≈ fck,cyl / 0.82 (cylinder is approximately 82% of cube). UK production testing traditionally used cubes; EC2 design uses cylinders. The TMS applies equally to both bases — just ensure tests and design use the same basis.
When a production source has fewer than 15 test results, BS EN 206 Table 14 requires using a fixed minimum margin rather than the statistical kσ approach: +4 N/mm² for fck < 35 N/mm², and +6 N/mm² for fck ≥ 35 N/mm². Once 15 or more results are available, the plant transitions to the statistical approach using the calculated σ from production data.
Conformity testing assesses whether a production run (typically 50+ m³) conforms to the specified class as a whole, using the mean of n results vs Criterion 1 and individual results vs Criterion 2. Identity testing is spot-checking of individual deliveries, using a simplified single-result check. TMS applies to conformity assessment; identity testing uses simplified criteria. Both are defined in BS EN 206 Clause 8.
When GGBS or PFA is used as a cement replacement (as in many durability-specified mixes), the 28-day strength is lower than for an equivalent CEM I mix at the same w/c ratio, but 90-day and later-age strength is higher. Mix design and TMS calculations for GGBS/PFA concretes should use the appropriate maturity age for strength assessment (often 56 days for GGBS-rich mixes) or the relevant equivalence factor from BS EN 206 Annex E.
Every ready-mix delivery to site must be accompanied by a delivery ticket (delivery note) specifying the concrete designation, strength class, w/c ratio, cement content, and admixtures. The delivery ticket is the primary documentary evidence that the specified concrete was supplied and that the TMS calculation has been applied. Building Control will request delivery tickets for structural works — always retain them.
BS EN 206 specifies minimum sampling rates for conformity testing. For continuous production: a minimum of one sample per 150 m³ or one per day of production, whichever gives more samples. For site-specific assessment: a minimum of one sample per 25 m³ for the first 50 m³, then one per 50 m³. Each sample typically comprises two 150mm cubes (or 150mm diameter cylinders): one tested at 7 days, one at 28 days.
If a concrete production lot fails the conformity criteria (mean result or individual result), BS EN 206 Clause 8.4 requires: investigation of the cause; a decision on the suspect concrete (which may include in-situ testing, core drilling, or structural assessment); and corrective action to prevent recurrence. Non-conforming concrete does not automatically mean demolition — in-situ core strength is typically 85–95% of standard cube strength, and structural engineers can assess whether the structure remains safe.
The fundamental relationship between w/c ratio and concrete strength is described by Abrams’ Law (1919): fc = A / Bw/c, where A and B are empirical constants. In practical terms, reducing the w/c ratio by 0.05 increases concrete strength by approximately 3–5 N/mm² for CEM I mixes in the typical structural range. This relationship is the basis for all w/c-based mix design methods.
UK concrete mix design methods include: the BRE/DoE method (Building Research Establishment, widely used in UK industry); the ACI 211 method (American approach, used internationally); and proprietary plant-specific methods. All start from TMS and use it to determine the w/c ratio via a strength-w/c relationship established from trial mixes and production data. TMS is always the starting point.
Target mean strength (TMS), also called required mean strength or aim strength, is the mean compressive strength that a concrete mix must be designed to achieve in production. It is always greater than the specified characteristic strength (fck) by a statistical margin M = kσ, where k is a factor related to the permitted proportion of low results (typically 1.64 for 5% defects) and σ is the standard deviation of the production process. TMS is the primary output of the mix design process and determines the required w/c ratio — a higher TMS requires a lower w/c ratio (more cement), increasing cost. For a C30/37 concrete with σ = 4 N/mm²: TMS = 30 + (1.64 × 4) = 36.6 N/mm².
Because concrete production has inherent variability — even the most carefully controlled ready-mix plant will produce batches with slightly varying strengths due to variations in aggregate moisture, mixing time, aggregate grading, cement batch variation, temperature, and human factors. If the mix were designed to achieve exactly fck as its mean strength, then by definition 50% of results would fall below fck — far more than the permitted 5%. By setting the mean strength (TMS) above fck by a margin proportional to the variability (σ), the proportion of results below fck is statistically limited to 5% (or whatever the specified defect rate is). The better the production control (lower σ), the smaller the margin needed and the closer TMS can be to fck.
For UK ready-mix concrete plants in 2026, typical production standard deviations (measured from 28-day cylinder or cube results) are: excellent control: σ = 2.5–3.5 N/mm² (continuous production, automated batching, high-frequency testing, experienced operators); good control: σ = 3.5–5.0 N/mm² (well-managed plant, regular production, adequate testing frequency — most established UK ready-mix plants); moderate control: σ = 5.0–7.0 N/mm² (less frequent production, variable aggregate supply, or limited quality systems); poor control: σ > 7 N/mm² (site-mixed or poorly controlled production — not acceptable for structural concrete). The BRMCA and BS EN 206 guide ready-mix plants to achieve and maintain σ below 5 N/mm² for continuous production.
BS EN 206 specifies two conformity criteria that must both be satisfied for a production lot to be considered conforming: Criterion 1 (mean): The mean of n consecutive test results must be ≥ fck + k1σ, where k1 = 1.48 for n ≥ 15 results, or the mean must be ≥ fck + 4 N/mm² for n = 3–14 results (initial production). Criterion 2 (individual): Every individual test result must be ≥ fck − 4 N/mm². Both criteria apply independently and simultaneously. A production lot that passes Criterion 1 (good mean) but has one very low individual result may still fail Criterion 2. Conversely, a lot with all individual results above fck but a low mean may fail Criterion 1 if the average is too low.
Both k = 1.48 and k = 1.64 correspond to the same 5% defect rate (5th percentile) from a normal distribution, but they apply to different aspects of the TMS and conformity process. k = 1.64 (or 1.645) is the exact value for the 5th percentile of a normal distribution and is used for mix design TMS calculation — setting the target the production process must aim for. k = 1.48 is used for conformity assessment (Criterion 1) in BS EN 206 — it is the value used to check whether a sample of n ≥ 15 results from production represents a conforming population. The difference between the two values reflects the additional uncertainty in estimating the true mean from a finite sample of n results rather than knowing the true population mean.
TMS has a direct impact on concrete production cost because a higher TMS requires a lower w/c ratio, which in turn requires more cement per m³. Approximately: each 1 N/mm² increase in TMS requires approximately a 0.005–0.01 reduction in w/c ratio; each 0.05 reduction in w/c ratio requires approximately 15–25 kg/m³ more cement; at current UK cement prices (approximately £120–£180/tonne in 2026), each extra 20 kg/m³ of cement adds approximately £2.40–£3.60/m³ to concrete cost. Therefore, a plant with σ = 6 N/mm² compared to σ = 4 N/mm² must target 3.3 N/mm² higher TMS, requiring approximately 10–15 kg/m³ more cement and adding approximately £1.50–£2.70/m³ extra cost per cubic metre of concrete produced. Over a year’s production at 100,000 m³/year, this represents £150,000–£270,000 in additional cement costs — a powerful economic incentive to invest in production control.
If test results from a batch or production lot fail either of the BS EN 206 conformity criteria, a structured investigation must follow. The steps typically are: (1) Check that the test procedure was followed correctly (correct sampling, curing, testing age, machine calibration) — many apparent failures are testing errors. (2) If the test is confirmed valid, identify the suspect concrete and place a hold on any decision until further assessment. (3) Commission in-situ testing of the hardened concrete in the structure — either rebound hammer tests (as a screening tool), pull-out tests (Lok test), or core drilling (most reliable). In-situ core strength is typically assessed against a modified criterion (approximately 0.85× the specified cube strength) to account for in-situ vs lab curing conditions. (4) A structural engineer assesses whether the structure is adequate given the measured in-situ strength. (5) In most cases, the structure is found adequate; genuinely inadequate concrete requiring demolition is rare but does occur. (6) Root cause analysis and corrective action to prevent recurrence.
In principle, yes — the TMS formula applies to any concrete production, including site-mixed. However, site-mixed concrete has much higher variability than ready-mix (σ typically 6–10 N/mm² or more due to inconsistent batching, varying aggregate moisture, and unreliable mixing) and is therefore inherently less economic for achieving a given characteristic strength — it requires a much higher TMS and therefore much more cement. For any structural application requiring fck ≥ 20 N/mm², ready-mix concrete from a BS EN 206-certified batching plant is strongly preferred over site-mixing. TMS calculations for site-mixed concrete should use a conservatively high σ (6–8 N/mm² minimum) unless actual production data from the specific site plant is available.
Additional TMS and mix design resources 2026: BSI – BS EN 206 / BS 8500 | The Concrete Society | The Concrete Centre | BRMCA | BRE Group – Mix Design Methods