Basis: a survey of 127 plants, not a national standard
The mean and 80th-percentile columns come from one study: Guo Hongli, Li Xinwei, Ren Qinyi, Liu Siyi, Wu Yunsheng and Jia Limin, “Characteristics of influent quality at typical municipal wastewater treatment plants in China”, Water & Wastewater Engineering, 2018(6). The sample is 127 municipal wastewater treatment plants in 19 provinces, using the 2016 annual average. It is a survey sample, not a national census. GB 50014-2021 does not publish a typical influent concentration table: it requires the design water quality to be established from a survey or from neighbouring towns, and §4.2.1 gives per-capita design loads when such data are absent. Industrial wastewater discharged to a sewer is governed separately by GB/T 31962-2015. For export projects serving up to 1,000 population, the loads are usually derived from the British Water per-person loads of 60 g BOD and 8 g ammonia-N per person per day, not from the concentrations below; above 1,000 population, use a municipal design basis.
Design influent quality
Values in mg/L. Table 1 reports the survey means; Figure 1 reports the P80 values. The two columns describe the same survey from two different angles, and neither is a design value.
| Parameter (mg/L) | Survey mean (127 plants, 19 provinces, 2016) | 80% of plants below (P80) |
|---|---|---|
| COD | 220 | 282 |
| BOD₅ | 81.6 | 105 |
| SS | 148.5 | 210.8 |
| NH₃-N | 22.8 | 27.6 |
| TN | 30.4 | 36.4 |
| TP | 3.7 | 3.3 |
Survey statistics, not a national-standard value. GB 50014-2021 does not give a typical influent concentration table; it requires the design water quality to be established from a survey or from neighbouring towns, and §4.2.1 gives per-capita design loads for use when data are absent.
What the mean, the 80th percentile and a design value each answer
The table gives the survey figures. This section explains the sample behind them, what the mean and the 80th percentile each describe, why the total-phosphorus mean is higher than its 80th percentile, what the national standard provides instead of a concentration table, and where the statistics stop applying. Value sources and explanation sources are kept separate.
The survey behind the two columns
127 plants, 19 provinces, 2016 annual averageSurvey literature
- What it is for
- The mean and 80th-percentile columns are two summaries of one 2016 survey of municipal wastewater treatment plants in 19 Chinese provinces.
- Why the value is stated this way
- Guo Hongli, Li Xinwei, Ren Qinyi, Liu Siyi, Wu Yunsheng and Jia Limin, “Characteristics of influent quality at typical municipal wastewater treatment plants in China”, Water & Wastewater Engineering, 2018(6). The paper reports 127 plants in 19 provinces; Section 3.4 reports the 2016 annual average used for the means in Table 1 and for the 80th percentiles in Figure 1.
- How to choose
- Use the figures as a planning reference for municipal sewage in China where no project analysis exists. Read both columns as descriptions of the same sample, not as two independent studies.
- When not to use it
- The sample is 127 plants in 19 provinces, not a national census and not a statutory national mean. It does not represent one city, one industrial catchment or one season.
Explanation basis: Definition of the sample
What the survey mean answers
COD 220 · BOD₅ 81.6 · SS 148.5 · NH₃-N 22.8 · TN 30.4 · TP 3.7 mg/LSurvey literature
- What it is for
- The mean answers how concentrated the influent is on average across the surveyed plants. It is an arithmetic average of the sample, used for a first indication of load.
- Why the value is stated this way
- The values are the sample means in Table 1 of the study. They describe the 127-plant sample; the page rounds the paper values to one decimal place where the paper gives more.
- How to choose
- Use the mean to estimate the average load of a plant similar to the sample. Do not read it as the concentration of a typical single plant or as a design value.
- When not to use it
- A mean is sensitive to a few extreme plants. It is not a maximum, a design concentration, a discharge standard or an equipment guarantee.
Explanation basis: Definition of an arithmetic mean
What the 80th percentile answers
COD 282 · BOD₅ 105 · SS 210.8 · NH₃-N 27.6 · TN 36.4 · TP 3.3 mg/LSurvey literature
- What it is for
- The 80th percentile answers a different question: below what concentration do 80% of the surveyed plants fall? It describes the upper part of the distribution, not its centre.
- Why the value is stated this way
- The values are read from Figure 1 of the study. The page keeps the paper values and rounds only where the paper gives more decimal places than the precision shown.
- How to choose
- Use the 80th percentile when the question is how much of the sample lies below a level, for example to judge how unusual a high-strength influent is. Do not substitute it for the mean.
- When not to use it
- The 80th percentile is not a removal target, not a compliance limit and not a design value. A plant above the 80th percentile is not outside a standard.
Explanation basis: Definition of a percentile
Why the TP mean (3.7) is above its 80th percentile (3.3)
TP range 0.32–36.52 mg/LSurvey literature
- What it is for
- A mean can exceed the 80th percentile when a distribution is right-skewed. This is a property of the sample, not a data error and not a reason to change either figure.
- Why the value is stated this way
- The study reports a TP range of 0.32–36.52 mg/L. The reported range, mean and 80th percentile are consistent with a right-skewed distribution: high values can pull the arithmetic mean up to 3.7 mg/L while 80% of the plants remain below 3.3 mg/L. This is an interpretation of the reported statistics, not a count of high-TP plants and not a causal attribution.
- How to choose
- If the project needs a central value, read the distribution and the 80th percentile; if it needs an average load, use the mean. State which one is being used.
- When not to use it
- The mean is not the typical plant just because it is an average, and the 80th percentile is not a corrected mean. Do not present either as a maximum or a design concentration.
Explanation basis: Definition and statistical interpretation of a right-skewed sample
What GB 50014-2021 provides instead of a concentration table
BOD₅ 40–60 · SS 40–70 · TN 8–12 · TP 0.9–2.5 g/(person·day)National standard clause
- What it is for
- The standard does not publish a table of typical influent concentrations. It requires the design influent quality to be established from a survey or from neighbouring towns, and §4.2.1 gives per-capita design loads when such data are absent.
- Why the value is stated this way
- GB 50014-2021 §4.2.1 gives per-capita design loads of BOD₅ 40–60 g/(person·day), SS 40–70 g/(person·day), TN 8–12 g/(person·day) and TP 0.9–2.5 g/(person·day). These are loads per person per day, not concentrations in mg/L, and they are converted with the design population and flow.
- How to choose
- Use the per-capita loads to build a design load when no survey is available, then combine them with the project population and flow. Use a measured survey where one exists.
- When not to use it
- These are design loads in g/(person·day), not influent concentrations, not the survey values in the table and not a discharge standard. Do not put them in an mg/L column or compare them directly with the table.
Value source:GB 50014-2021, Code for design of outdoor wastewater engineering
Explanation source:GB 50014-2021, Code for design of outdoor wastewater engineering
Measured data come first
Measured project data over planning statisticsNational standard requirement
- What it is for
- A planning chart cannot replace a project water analysis. The standard requires the design water quality to be based on survey data or on neighbouring towns, so a measured project sample is the first basis for the design.
- Why the value is stated this way
- GB 50014-2021 §4.2.1 requires the design water quality to be established from a survey or from neighbouring towns; the per-person design loads are the fallback when such data are absent. A project with its own sampling should design on that sample, including its seasonal and industrial variation.
- How to choose
- Sample the actual influent, state the sampling period and flow basis, and use the measured result for the design. Keep this chart only as a cross-check.
- When not to use it
- Do not use this chart to override a measured result, and do not present a chart value as a guarantee of the influent a plant will receive.
Value source:GB 50014-2021, Code for design of outdoor wastewater engineering
Explanation source:GB 50014-2021, Code for design of outdoor wastewater engineering
Industrial wastewater and sewer discharge
GB/T 31962-2015National standard
- What it is for
- Industrial wastewater, or sewage with a high industrial fraction, has a different composition from the municipal sample and must be sampled separately before a treatment train is fixed.
- Why the value is stated this way
- GB/T 31962-2015 is the water-quality standard for sewage discharged into urban sewers. It controls what an industrial discharger may put into the municipal sewer; it is not an influent design table for the treatment plant.
- How to choose
- Identify the industrial connections, obtain their discharge data and sample the combined influent. Check the sewer-discharge standard separately from the treatment design.
- When not to use it
- Do not apply the municipal survey statistics to an industrial or high-industrial-fraction influent, and do not read GB/T 31962-2015 as an effluent standard for the treatment plant.
Export projects: per-person loads
60 g BOD · 8 g ammonia-N (as N) per person per day (serving up to 1,000 population)British Water code of practice
- What it is for
- For export projects serving up to 1,000 population, the plant is often sized from the population served and a per-person load rather than from a measured influent concentration.
- Why the value is stated this way
- British Water, Code of Practice — Flows and Loads 4 (BW COP 18.11/13, revised 2013), domestic dwellings, standard residential row: 60 g BOD and 8 g ammonia-N (as N) per person per day. The code’s residential figures apply to systems serving up to 1,000 population.
- How to choose
- For an export project serving up to 1,000 population, derive the daily load from the design population and these per-person figures, then size the process on that load and the project flow. For a larger population, use a municipal design basis instead.
- When not to use it
- These are per-person loads for small systems serving up to 1,000 population, not concentrations. Do not mix them with the mg/L survey columns, and do not apply them above 1,000 population without a municipal design basis.
What these statistics are not
Not a discharge standard, an equipment guarantee or a statutory design valueDefinition and boundary
- What it is for
- A design value answers which concentration the process must be sized for. The columns describe a sample; they do not set a legal limit, a supplier’s performance guarantee or the design value that a standard requires.
- Why the value is stated this way
- A discharge standard is set by the environmental authority and applies to the treated effluent; an equipment guarantee is a contractual commitment for a defined influent and acceptance method; a statutory design value is a value a design code requires. A survey statistic is none of these.
- How to choose
- Use the statistics to frame a planning question, then fix the design from the project survey, the applicable standard and the contract documents. State the role of each number.
- When not to use it
- Do not quote a survey mean or percentile as a compliance limit, a guaranteed effluent or a design value. Standards and contracts are separate sources with their own conditions.
Value source:Definition of a statistical mean, a percentile and a design valueGB 50014-2021, Code for design of outdoor wastewater engineering
Explanation basis: Definition and engineering interpretation
Sources
Guo Hongli, Li Xinwei, Ren Qinyi, Liu Siyi, Wu Yunsheng and Jia Limin. Characteristics of influent quality at typical municipal wastewater treatment plants in China. Water & Wastewater Engineering, 2018(6).
Sample section (127 plants, 19 provinces); Section 3.4 (2016 annual average); Table 1 (means); Figure 1 (80th percentiles); Section 3.2 text (TP range 0.32–36.52 mg/L).
Full-text reprint used for verificationGB 50014-2021, Code for design of outdoor wastewater engineering
§4.2.1 (design water quality by survey or neighbouring towns; per-capita design loads: BOD₅ 40–60, SS 40–70, TN 8–12, TP 0.9–2.5 g/(person·day)).
Full-text mirror used for cross-checkBritish Water. Code of Practice — Flows and Loads 4 (BW COP 18.11/13, revised 2013).
Domestic dwellings, standard residential row: 60 g BOD and 8 g ammonia-N (as N) per person per day; systems serving up to 1,000 population.
Official code of practice PDFGB/T 31962-2015, Water quality standard for sewage discharged into urban sewers
Scope and current status (issued 2015-09-11, effective 2016-08-01).
National standards platform recordDefinition of a statistical mean, a percentile and a design value
Terminology used on this chart; no external numeric source.
How to use and notes
- The figures are survey statistics, not a national-standard value. They come from a 2016 survey of 127 municipal wastewater treatment plants in 19 provinces; the mean and the 80th percentile are from the same study.
- Where a project has measured influent data, use the measured data; this chart is a planning reference only.
- The mean is the average of the surveyed plants; the 80th percentile is the value below which 80% of those plants fall. They answer different questions and must not be interchanged. The TP mean (3.7 mg/L) is above its 80th percentile (3.3 mg/L); the reported range of 0.32–36.52 mg/L indicates a right-skewed distribution, so the arithmetic mean is pulled above the 80th percentile. This is an interpretation of the reported statistics, not a count of high-TP plants, and the mean is not a typical-plant value.
- Industrial wastewater, or sewage with a high industrial fraction, needs its own sampling; GB/T 31962-2015 governs discharge to a municipal sewer.
- For export projects serving up to 1,000 population, organic and ammonia loads are usually derived from per-person loads (60 g BOD and 8 g ammonia-N per person per day) rather than from the concentrations above; above 1,000 population, use a municipal design basis, and do not mix either method with the mg/L columns.
- The influent range our equipment accepts and the pretreatment required outside that range are confirmed for each project from its own water analysis and process design.

