top of page

Hydronic Flushing Flow: Check Every Pipe Section Before Choosing a Temporary Pump

Oct 16
5 min read
Original teal and cream engineering illustration of a main splitting into two pipe branches, emphasizing section-by-section flushing flow checks.

A flushing pump can produce the scheduled total flow while part of a chilled-water network receives too little movement to satisfy the cleaning plan. The common explanation is a missing section check: the pump flow, branch flows and pipe areas have been treated as interchangeable. A second problem appears when a pump selected for normal operation is expected to deliver a temporary flushing duty through different hoses, bypasses and filtration equipment.


Prepare the flushing calculation around a marked flow route. The useful output is a schedule of actual bores, required section flows, permitted valve arrangements and measured evidence. Pump selection follows that schedule. This article develops an original hydraulic example; it does not prescribe a cleaning velocity, chemical treatment or acceptance standard.


1. Agree the cleaning basis before calculating flow


The cleaning specialist and project team should establish the applicable method, section-specific velocity criteria, equipment protection, water-quality acceptance and records. The CIBSE-hosted closed-system water-treatment presentation, particularly pages 7 and 24–27, brings together design information, sensitive equipment, bypasses and flushing-flow considerations. Its tabulated guidance must be read in the context of the relevant project specification and cleaning method.


Do not select one convenient velocity from a presentation and apply it indiscriminately. Pipe material, internal diameter, normal design duty, permitted equipment velocity and the intended removal process all need consideration. Confirm which components remain in the route and which need approved bypass arrangements. A calculation that succeeds only by sending debris through a protected coil is not a usable plan.


For a circular pipe, required volume flow is V = v × πD²/4. Use internal diameter D in metres and velocity v in m/s to obtain m³/s; multiply by 1,000 for L/s. KSB's head-loss reference defines this velocity-and-bore relationship and the accompanying pipe-resistance terms. Nominal pipe size alone is insufficient.


2. Test an original main-and-branch arrangement


Assume a hypothetical flushing plan specifies 1.20 m/s for three particular pipe sections. This common target is invented to expose the flow accounting and is not a recommended minimum. The main has an actual bore of 100 mm. It divides into branches with actual bores of 80 mm and 50 mm, which subsequently rejoin. Ignore minor density changes and assume all three sections are full of water.


The 80 mm branch needs 1.20 × π × 0.080²/4 × 1,000 = 6.032 L/s. The 50 mm branch needs 1.20 × π × 0.050²/4 × 1,000 = 2.356 L/s. Together they carry 8.388 L/s. These are required target flows; they are not a prediction that opening both branches will produce that split.


At the sum of those branch targets, main velocity is 0.008388 ÷ (π × 0.100²/4) = 1.068 m/s. Both branches can therefore meet their individual target while the main falls below its assumed 1.20 m/s requirement. The main itself needs 9.425 L/s, leaving a difference of 1.037 L/s above the sum of the branch targets.


One possible allocation is 7.069 L/s through the 80 mm branch and 2.356 L/s through the 50 mm branch. The corresponding velocities are 1.406 and 1.200 m/s, with 1.200 m/s in the main. Whether that arrangement is acceptable depends on the approved maximum velocities, available pressure and actual hydraulic balancing. It is a candidate for evaluation, not an instruction to adjust valves to those numbers.


Original flow example: 80 mm and 50 mm-bore branch targets of 6.032 and 2.356 litres per second total 8.388, producing only 1.068 metres per second in a 100 mm main. The assumed main target requires 9.425 litres per second.

Original example using an assumed 1.20 m/s target, not a prescribed cleaning velocity. Branch target flows alone leave the common main short.


3. Check each stage, including isolated branches


Suppose the two branches are instead flushed separately. Passing 6.032 L/s through the 80 mm branch gives only 0.768 m/s in the common main. Passing 2.356 L/s through the 50 mm branch gives 0.300 m/s in that main. Recording both branch stages does not establish that the main ever received its own required flow.


Forcing the full 9.425 L/s main target through the 50 mm branch would raise its velocity to 4.800 m/s. The arithmetic is straightforward, but that does not establish that the branch, its fittings or connected equipment can tolerate the condition. Review a separate main-flushing stage, a properly sized temporary connection or another specialist-approved method when the geometry creates this conflict.


The general lesson is to mark every stage on the schematic and calculate every section traversed in that stage. Include common returns and temporary links. A record headed only “pump flow achieved” cannot distinguish an adequately flushed main from an inadequately served branch. For the related operating-network principle, see parallel hydronic branches and resistance.


4. Build the temporary-pump duty on its actual route


Consider a separate main-flushing configuration with sensitive branches bypassed under the approved plan. At 9.425 L/s, assume the entire circulating route contains 60 m of 100 mm-bore pipe and a total of 20 m of 80 mm-bore temporary hose. Assume water density 1,000 kg/m³, g = 9.81 m/s² and a Darcy friction factor of 0.025 for both materials. The fixed friction factor is a teaching assumption; obtain actual hose and pipe data for a real selection.


The pipe velocity is 1.200 m/s and hose velocity is 1.875 m/s. Using Δp = f(L/D)ρv²/2, straight-pipe loss is 10.800 kPa and straight-hose loss is 10.986 kPa. Assume local-loss coefficients total 6.0 referenced to the 100 mm bore, and 5.0 referenced to the 80 mm hose bore. Their losses are 4.320 and 8.789 kPa respectively.


Add an explicitly assumed 15.000 kPa clean-filtration loss at this flow. Total route loss is 10.800 + 10.986 + 4.320 + 8.789 + 15.000 = 49.895 kPa. Required differential head is 49,895 ÷ (1,000 × 9.81) = 5.086 m. The resulting screening duty is 9.425 L/s at 5.086 m, before any justified project allowances or changed filtration condition.


This example assumes a completely filled recirculating route returning to the same pressure reference, with no net static lift. A once-through discharge, different vessel levels or different terminal pressures would require the corresponding static terms. Check suction conditions, hose pressure ratings, restraints, flow-measurement arrangements and the filter's permitted loading. A clean-only calculation does not establish performance throughout the cleaning cycle.


5. Verify flow and cleanliness separately


  • Record the approved flushing stage, valve lineup, actual pipe bore and target flow for each section.

  • Confirm measuring-device suitability, installation requirements, calibration and enough measurement resolution to judge the target.

  • Log stable section flows and pump differential pressure with the filter condition and pump speed.

  • Keep the specified sampling, inspection and water-quality acceptance results alongside hydraulic records.

  • Record reinstatement of equipment, removal of temporary links and the final operating configuration.


Hydraulic evidence establishes what flow was achieved. It cannot by itself prove that debris, chemical residues or microbiological conditions satisfy the cleaning specification. Where either hydraulic or cleanliness acceptance fails, return the issue to the responsible project team rather than treating elapsed flushing time as proof of completion.


For the underlying pipework calculations, the paid MEP Pipe Design and Sizing Handbook helps organize flow derivation, pipe dimensions, resistance and pump-head review. Use that broader design workflow to make the hydraulic schedule easier to check, alongside the specialist's project-specific cleaning method and acceptance evidence.



Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
bottom of page