Engineering Knowledge Cards.
Confidence HighUpdated 2026-11-01

The essential difference between flow-line construction (the flow method of organising work) and network planning techniques, an element-by-element comparison of the two systems (principle / presentation tool / optimisation objective / applicable scenario) and the ways the two are combined (GB/T 50502-2009 clauses 5.2.2, 6.2.2, 6.3.1; JGJ/T 121-2015 clauses 3.2.14, 3.2.17, 4.5.1, 6.2 to 6.4)

Flow-line construction is a method of organising work: it decides how space is divided (flow sections), how crews move (rhythm and flow-line spacing), and how resources are levelled. Network planning is a technique for analysing programmes: it decides how activities are linked (logic and time lags), how times are computed (ES/EF/LS/LF, TF/FF), and how float is used (duration optimisation, resource-limited and resource-levelling optimisation, time-cost optimisation). The two are not competing methods of the same kind; they are two layers - spatial rhythm and the time skeleton - and the standards link them in sequence: GB/T 50502-2009 clause 6.2.2 ('the sequence of construction and the flow sections shall be fixed in the construction arrangements') and clause 6.3.1 (the progress programme 'shall show the lapping relationships between construction sections or trades, the durations, and the start and finish times') - that is, build the network plan on flow sections as units, then optimise using network float; JGJ/T 121-2015 clause 8.2.3 permits conversion between network diagrams and bar charts, allowing interface with the flow-line bar chart.

Applicable Codes

  • GB/T 50502-2009 · Code for Construction Organization Planning of Building Works 第 5.2.2、6.2.2、6.3.1 条
  • JGJ/T 121-2015 · Specification for Engineering Network Planning Technology 第 3.2.14、3.2.17、4.4.1、4.5.1、5.4.1、6.2.2、6.3.1、6.3.4、6.4.2、8.2.3 条

Topic Scope

This card compares the two families of tools for construction progress control - flow-line construction (the flow method) and network planning techniques (JGJ/T 121-2015) - giving an element-by-element comparison of the two systems (principle / presentation tool / optimisation objective / applicable scenario), their normative linkage within the project execution plan and the progress programme, and three ways of using them together. It answers 'which system fits a given scenario', 'can the two be stacked', and, when stacked, what comes first and what comes after.

Core Conclusions

1. One-line characterisation

Flow-line construction answers 'how space is divided, how crews move, how resources are levelled'; network planning answers 'how activities are linked, how times are computed, how float is used'. The former is a method of organising work (trades flowing horizontally across sections); the latter is a technique for analysing programmes (time computation and optimisation of activity logic). The two are not rivals of the same kind but two layers - 'spatial rhythm' and 'the time skeleton' - and can, and should, be used together.

2. Element-by-element comparison table

Element Flow-line construction (flow method) Network planning techniques (JGJ/T 121-2015)
Disciplinary position A method of work organisation within construction organisation An analysis technique of programme management, governed end to end by JGJ/T 121-2015
Basic principle The works are divided in plan or vertically into construction sections; specialist crews advance section by section in a fixed order at fixed (or multiple) rhythms; adjacent crews keep the minimum flow-line spacing K, so that several sections run in parallel in the same period while each crew works continuously, achieving continuous, level, rhythmic production Arrows plus nodes (double-code, i.e. activity-on-arrow) or nodes alone (single-code, i.e. activity-on-node) form a directed network expressing the sequence logic of activities; earliest and latest times, total float and free float are computed; the path of longest total duration fixes the critical path, and float is treated as a disposable reserve of time
Core parameter system Number of construction processes n, number of sections m, flow rhythm t, flow-line spacing K, technical and organisational breaks sum(Z), lapping time sum(C); duration T = sum(K) + the sum of the rhythms of the last process over its sections + sum(Z) - sum(C) Activity duration D, logic relations and time lags (FTS/STS/FTF/STF), nodes, paths; calculated duration Tc, required duration Tr, planned duration Tp; total float TF, free float FF, time interval; earliest and latest node times
Presentation tools Bar chart (Gantt chart), flow-line progress table (sloped-line chart), dynamic curves of labour (resource) demand Double-code network diagram, single-code network diagram, double-code time-scaled network diagram, single-code lapped-network diagram, multi-level (hierarchical) network plans
Computational outputs Crew start times on site, start and finish of work on each section, flow duration, resource demand curves ES/EF/LS/LF and TF/FF of each activity, critical activities and the critical path, calculated duration, resource demand curves
Optimisation objective Levelling-oriented: aims at continuous, level, rhythmic production, reducing idle time and trade interference; for given m, n and t seeks the shortest duration, or trades duration by adding crews or sections or reducing K (in effect a duration-for-resource swap) Metric-oriented: guided by one measure (duration, resource or cost), uses float adjustment to keep improving the initial plan toward an optimum - duration optimisation, resource-limited optimisation, resource-levelling optimisation, time-cost optimisation
Optimisation means Adjust m (add or remove sections), adjust n (merge or split processes), adjust t (add resources), add specialist crews for multiple rhythms, adjust K and lapping Compress critical activities (six-step duration optimisation), the 'limited resources, shortest duration' sequencing, the 'fixed duration, level resources' peak-shaving method, time-cost optimisation compressing by direct-cost slope
Decision granularity Working level / crew level, monthly and weekly work plans Management level / project level: master control programmes, milestones, EOT claims and their demonstration
Applicable scenarios Works of regular structure that can be sectioned into equal parts with strongly repetitive trades: standard floors of multi-storey housing, linear road and pipeline works, rows of foundations and piers, staged tunnel driving, standard-room fit-out Many trade interfaces and complex logic; a need to identify the critical path, use float, optimise resources and cost, or argue EOT claims; also multi-level plans (master network plus sub-networks)
Main limits Hard to express non-flow logic directly (such as 'start only after all of the preceding is complete'); gives no float or critical path directly; resource peaks need separate counting Does not itself carry the spatial intent of level production; too fine a work breakdown swells the diagram; without section-based modelling it cannot show crew continuity

3. What the standards say about the linkage (key wording)

GB/T 50502-2009, the national code for project execution plans of building works

  • Clause 5.2.2: progress arrangement and spatial organisation within the construction deployment shall comply with the following: '1 the main construction content and its progress arrangement shall be stated clearly, and the construction sequence shall accord with trade logic; 2 flow sections shall be divided by stages in light of the specifics of the works; the stages in the execution plan for a building work generally include three: foundation and substructure, main structure, and fit-out and M&E installation.' The clause also provides: 'the construction deployment shall make overall arrangements for the main divisional (sub-divisional) works and specialist works of this building and state the milestone nodes of the construction process', and 'the division of flow sections shall be reasonable in light of the character and quantities of the works, the basis and the direction of flow shall be stated, and level flow construction shall be ensured'.
  • Clause 6.2.2: 'The sequence of construction and the flow sections shall be fixed in the construction arrangements.'
  • Clause 6.3.1: 'The progress programme for divisional (sub-divisional) works or specialist works shall be compiled under the construction arrangements and in conjunction with the general contractor's progress programme.' 'The compilation shall be comprehensive in content, reasonable in arrangement, scientific and practical; the programme shall show the lapping relationships between construction sections or trades, the durations, and the start and finish times. The programme shall also embody and implement the target-control requirements of the overall programme; compiling the divisional (sub-divisional) or specialist programme in turn demonstrates the soundness of the overall programme.'

JGJ/T 121-2015, the specification for network planning techniques in engineering

  • Clause 3.2.14: the time parameters of a network plan shall include 'the earliest start, earliest finish, latest start and latest finish times of activities, total float and free float; earliest and latest node times; time intervals; calculated duration, required duration and planned duration', and 'the time parameters of a network plan should be computed with computer software'.
  • Clause 3.2.17: the explanatory notes of a network plan shall include '1) compilation notes; 2) a schedule of the main plan indicators; 3) key notes for implementing the plan; 4) problems to be solved and the main measures; 5) the scope of allocation of float among activities; 6) other matters requiring explanation'; 'the formal network plan to be implemented shall be produced on the basis of the optimisation results and submitted for approval.' Optimisation is defined: 'network plan optimisation is the process of, under given conditions and against one measure (duration, resource or cost), using float adjustment to keep improving the initial plan of the network toward an optimum. By the measure adopted it falls into duration optimisation, resource-limited optimisation, resource-levelling optimisation and time-cost optimisation.'
  • Clause 4.5.1: 'activities with the least total float are critical activities'; 'a path composed throughout of critical activities, or the path whose activity durations sum to the longest, is the critical path, which should be marked in bold, double or coloured lines'; where not all time parameters need computing, the node-labelling method can quickly fix the calculated duration and the critical path.
  • Clause 5.4.1: in a single-code lapped network the time lags are noted beside the arrows; the four basic lags are finish-to-start (FTS), start-to-start (STS), finish-to-finish (FTF) and start-to-finish (STF); 'the duration of a lapped network plan does not necessarily depend on the finish of the activities tied to the end node, but may depend on the finish of intermediate activities'.
  • Clause 4.4.1: a double-code time-scaled network plan 'shall show activities with solid arrows, dummy activities with dashed arrows, and the free float of activities with wave arrows'; the time unit may be the hour, day, week, ten-day period, month, quarter or year; the horizontal projection of every symbol on the time scale must correspond to its time parameters, and node centres must align with the corresponding scale positions.
  • Clause 6.2.2: six steps of duration optimisation: (1) compute and find the calculated duration, the critical activities and the critical path of the initial network plan; (2) compute the time to be shortened against the required duration; (3) determine the durational shortening available for each critical activity; (4) select critical activities to compress per the rules and recompute the calculated duration - 'where a compressed critical activity has turned non-critical, its duration shall be extended so that it remains critical'; (5) if the required duration is still exceeded, repeat steps 1 to 4; (6) if every critical activity is at its shortening limit yet the requirement is still unmet, revise the technical or organisational scheme or re-examine the required duration.
  • Clause 6.3.1: 'Network plans should be resource-optimised on the lines of limited resources with the shortest duration, and fixed duration with level resources.'
  • Clause 6.3.4: 'fixed duration, level resources' optimisation may use the peak-shaving method, using float to lower the resource peak. Steps: (1) compute the resource demand per time unit; (2) set the shaving target as the maximum per-unit resource demand minus one unit of resource; (3) find the last time of the peak period T_h and the earliest start ES and total float TF of the activities concerned; (4) compute the time difference dT (double-code dT_i-j = TF_i-j - (T_h - ES_i-j); single-code dT_i = TF_i - (T_h - ES_i)), prefer as the adjustment object the activity with the largest time difference and set ES = T_h; (5) when the peak can be reduced no further the optimised plan is reached; otherwise repeat steps 1 to 4.
  • Clause 6.4.2: time-cost optimisation screens by the direct-cost slope dC = (CC - CN) / (DN - DC) (CC the direct cost after compressing to the shortest duration, CN the direct cost under normal conditions, DC the shortest duration, DN the normal duration); with one critical path, compress the critical activity with the smallest slope; with several critical paths, compress the group of critical activities with the smallest combined slope; where the slope of the compressed object is no greater than the indirect-cost slope of the works, compression reduces or holds total cost, so compress; where it is greater, stop, and the preceding plan is the optimum. The shortening must satisfy 'not below the shortest duration' and 'a critical activity must not be compressed into a non-critical one'.
  • Clause 8.2.3: computer software should offer 'mutual conversion between single-code, double-code and time-scaled network diagrams', turning a network plan into a bar-chart programme drawn at earliest or latest times; for a time-scaled network with a long duration and widely differing activity durations, a non-uniform time scale may be used.

4. Three ways of combining them

  1. Sequential linkage in compilation (space first, time lands): complete the spatial organisation in the construction deployment under GB/T 50502-2009 clause 5.2.2 - divide the flow sections, state the basis and direction of flow, fix the construction sequence and the milestones; then compile the progress programme under clause 6.3.1. Clause 6.2.2 ('the sequence of construction and the flow sections shall be fixed in the construction arrangements') makes the flow-section conclusions an input to the progress programme.
  2. Mutual conversion of models (double networks embedded with flow sections): build the double-code or single-code network with flow sections as the modelling units, using STS, FTS and the other lags to express the lapping and minimum spacing of adjacent crews between sections - the network expression of flow-line spacing; in reverse, per JGJ/T 121-2015 clause 8.2.3, convert the network plan into a bar chart to interface with the flow-line progress table, keeping the work breakdown from becoming so fine that crew continuity disappears from view.
  3. Complementary optimisation objectives (constraining network optimisation with flow logic): flow-line construction seeks level resources, matching the 'fixed duration, level resources' peak-shaving of clauses 6.3.1 and 6.3.4, using float to shave peaks toward the average; it also seeks a shorter duration, matching the duration optimisation of clause 6.2.2, but compressing critical activities must respect paragraph 4 of clause 6.2.2 ('where a compressed critical activity has turned non-critical, its duration shall be extended so that it remains critical'), otherwise crew continuity is broken and new idle time created; where cost trade-offs arise, close with clause 6.4.2 by comparing the direct-cost slope with the indirect-cost slope.

Applicable Boundaries

  • This card positions the methods and frames the clauses; it does not replace the calculation of section division, flow duration or network parameters for a specific project - those follow the flow-type determination and duration formulae and the recursive formulae for network time parameters.
  • 'Flow-line construction' and 'division of flow sections' are not the same level of question: this card compares whole method families, and division principles are one sub-question within them.
  • Each of the four network optimisations has preconditions: duration optimisation needs compressible critical slack; resource levelling presupposes a fixed duration; time-cost optimisation presupposes quantifiable direct-cost and indirect-cost slopes, and does not apply where the indirect slope cannot be fixed.
  • The symbol rules for time-scaled networks apply only to double-code time-scaled networks (JGJ/T 121-2015 clause 4.4.1); the duration of a single-code lapped network is not necessarily set by the activities tied to the end node, and the double-code end-node rule must not be applied to it.
  • Clause numbers follow the editions cited; JGJ/T 121-2015 and GB/T 50502-2009 are both current recommendatory standards; where the construction contract or a local standard imposes stricter requirements, those prevail.

Further Reading

  • GB/T 50502-2009, the code for project execution plans of building works: chapter 5, execution plan for an individual building: 5.2.2 (progress arrangement and spatial organisation in the deployment), 6.2.2 (sequence and flow sections fixed in the arrangements), 6.3.1 (basis and content requirements of the progress programme).
  • JGJ/T 121-2015, specification for network planning techniques: chapter 3: 3.2.14 (list of time parameters), 3.2.17 (explanatory notes and the definition of optimisation); chapter 4: 4.4.1 (double-code time-scaled format), 4.5.1 (critical activities and path, node labelling); chapter 5: 5.4.1 (single-code lags); chapter 6: 6.2.2 (duration optimisation steps), 6.3.1 (two modes of resource optimisation), 6.3.4 (peak-shaving steps and formula), 6.4.2 (time-cost optimisation and the direct-cost slope); chapter 8: 8.2.3 (network-to-bar-chart conversion, non-uniform scale).
  • Next steps: flow-type determination (even rhythm / disparate rhythm / multiple rhythm / non-rhythmic) and flow-duration calculation; recursive formulae for double-code parameters by the activity method; reading ES/EF/FF directly off a time-scaled network; quantitative rules for combining the critical path method with flow-line construction to compress total duration.

Keywords

  • 流水施工
  • 网络计划技术
  • 施工流水段
  • 流水步距
  • critical path
  • 网络计划优化
  • 工期优化
  • 横道图
  • 施工进度计划
  • JGJ/T 121-2015