Frequently asked questions.

The suspect section is isolated and put under a defined pressure; over the holding time we record whether and how fast it drops. That separates the tight pipe from the leaking one before the actual leak search begins. Temperature changes in the pipe shift the pressure even without a leak – which is why it is recorded over time and not decided from one glance at the gauge.

Screed, plaster and insulation layer are scanned with moisture meters and the readings compared across the area. That shows how far the dampness reaches and where it is greatest. A single reading says little: salts in the masonry, residual moisture from construction and condensation produce similar readings. Only the pattern across many points gives a picture.

A thermal imaging camera shows temperature differences at the surface; a hot water leak stands out as a plume. The camera sees temperature, not water – insulation layers, sunshine and underfloor heating produce similar patterns. Thermography narrows things down; on its own it proves nothing.

The pipe is isolated, drained and filled with a test gas. The gas escapes at the defect, travels upwards through screed or ground and is indicated at the surface; dense floor coverings can divert it sideways.

A camera is guided through the wastewater pipe and records the run. Cracks, fragments, roots, open joints and backwater can be proven on the picture, with the metre reading from the entry point. At the surface, damage can be marked directly by means of a built-in sonde. Prerequisites are an accessible entry point and a flushed pipe – on a coated pipe wall the camera does not see the damage.

Through a small drill hole a probe is guided into the cavity: pre-wall installation, shaft, suspended ceiling, screed void. That replaces breaking open at the suspect point – what lies behind the building element is seen, with an intervention of a few millimetres.

A dye is added to the water and the suspect run is flushed. If the dye emerges at the damp spot, the path is proven: that separates wastewater from drinking water and one run from the neighbouring one. The method shows the connection, not the millimetre – the exact point is delivered by the locating methods.

An isolated area is filled and the water level is read over time – at manholes, basins and standing pipe sections. If the level falls, the section is leaking, and the drop per hour says how large the escape is. Evaporation on open surfaces is offset, otherwise you are measuring the weather.

Before the search it has to be clear where the pipe runs: a locating signal is applied to metallic pipes and traced at the surface, while a sonde is pushed into non-metallic pipes. Pipes running in parallel and reinforcement carry the signal across, which is why cross-checks are made at several points.

The waterproofing membrane is an electrical insulator, while the substrate beneath it is conductive: if the wetted surface is placed under voltage, current flows only where the waterproofing has a hole, and the measuring electrodes show the direction to the defect. In this way every single hole is found, including under ballast, as soon as the test field has been exposed. With metal-faced membranes, a missing conductive substrate or insulation that is already saturated, the method can only be used to a limited extent.

A demarcated roof segment is dammed up and the water level observed over a set period; evaporation and rainfall are taken out of the calculation by means of a reference vessel filled at the same time and placed in the test field. If the segment falls behind the reference vessel, it is leaking – the result, however, only says whether a segment is leaking, not at which point, and the structural capacity and the discharge of the water have to be settled beforehand.

A dye is added to the water and the suspect area is charged with it. If the dye emerges inside, the water path is proven and not merely assumed; in gravel-covered or planted build-ups the dye is heavily diluted.

The charging is done in levels, from the bottom upwards. We start at floor level; if nothing shows there, the next level up is watered. In doing so we deliberately use several dyes side by side to delimit the areas against one another – at the point of emergence it can then be read off which area carries the water. In this way we work our way up level by level until a result shows.

The method takes time, and not ours but the water’s: the dyed discharge often only shows in the room below after one to four days, because the water first has to find its way through the build-up. On the day of application there is therefore regularly nothing to be seen yet. A dye test consequently includes a follow-up check inside – not as an additional service, but because without the second look the method produces no result. We arrange the check appointment at the first visit. Check measurements are charged separately according to time spent and travel.

Saturated insulation stores heat differently from dry insulation: after sunshine a wet zone cools down more slowly and stands out in the thermal image. The method shows the extent of the wetting, not the hole, and needs a sufficient temperature difference.

Measuring is done above all inside: in the room beneath the roof we go over the traces of the water ingress and record how damp they are. That distinguishes whether an ingress is still fresh or dried out long ago – you cannot tell that from the stain on the ceiling.

Before anything else comes the preliminary check: is the heating losing water at all, or is something else behind it? The more precisely that has been settled in advance, the better. Where the suspect circuit is already known, we test in a targeted way and reach a result more quickly.

Testing is only possible on a system that has cooled down and is at room temperature – and that applies even when the circuit has already been narrowed down. If the system is still warm at the appointment, we flush it through on site; depending on the size of the heating system that takes several hours. Only after that is it prepared. Each circuit is then tested individually.

On request we take on the preliminary check ourselves. It then needs several check appointments at which the system is observed, before the pressure test proper and the search for the leak begin.

The circuit is drained and filled with a test gas: its hydrogen content escapes even at the smallest defects, rises through screed and floor covering and is indicated at the surface by the gas detector; in the mixture used, the gas is not ignitable. Dense floor coverings and membranes deflect it sideways, so that the point of emergence can lie next to the defect.

If a single circuit is run warm while the others stay isolated, its route stands out in the thermal image, and a point of escape often shows as a zone of differing warmth. With thick screeds or carpet the route can only be made out faintly.

At the heating manifold each circuit is isolated individually and it is observed which one does not hold the pressure. This step costs little and reduces the area to be examined from the whole flat to a single room; it requires an accessible manifold with functioning isolating valves.

A container filled with pool water is placed floating in the pool or on the pool edge; it evaporates under the same weather conditions as the pool surface but loses no water through leakage. The difference between the level in the container and the level in the pool over the same period is the actual loss – and incidentally it shows at what height the level comes to rest. Wind, bathing use, backwashing and rain distort the comparison, which is why use of the pool during the measurement period is recorded.

The good thing about this comparison: it needs no instrument and no technician, only a container, a sheet of paper and a few days of patience – which is why it is best left with you, and we are glad to explain on the telephone how it is done. It is not obligatory; we can narrow the loss down on site as well. But if your notes are already there, the visit does not begin with the question of whether the pool is leaking at all but straight away with the search for the point – that shortens the appointment noticeably, and you see the time saved directly on the invoice.

A bell built for exactly this purpose is placed over the suspect point – a liner weld, an inlet, the skimmer, a joint – and a negative pressure is built up beneath it. If the vacuum holds, the component is sound. If it falls away, the point is drawing air and the waterproofing is open there.

That is the difference from a visual inspection: the method assesses no area and weighs up no suspicion, it tests a single component and answers yes or no. For that the bell needs an accessible, exposed point – which is why the suspicion decides whether the pool is emptied for the test.

For the same reason, repeating it after the repair is not an extra but the second part of the same method: only that shows, at the same component, whether the point has become sound.

Individual pipe runs are isolated in the plant chamber and put under pressure. If one run holds the pressure and another does not, the leaking pipe has been identified without a single square metre of paving having been opened up. Solvent-welded collector pipes without a separation point cannot be assessed individually.

An ultra-flexible camera for pipe diameters from DN 30 to DN 70 is pushed into the inlet pipes and the rest of the pipework. It is built for exactly those tight cross-sections and bends at which an ordinary drainage camera can go no further. With it we see a defect that the pressure test or the vacuum bell has narrowed down directly in the pipe – a crack, an open socket, a solvent-weld joint that has come apart. Confirmation with our own eyes takes the last residual risk out of the repair: opening up only happens once it has been established what lies there.

A locating sonde additionally sits in the camera head. With the camera standing at the damaged point, its signal is taken as a bearing at the surface – the marking therefore sits exactly above the image that shows the defect, not above a calculated distance. The route of the pipe for the pipework sketch is run along in the same way. A separate locating run with a separate sonde is thereby unnecessary; only reinforcement in concrete can pull the bearing off line.

Two measuring principles complement one another: capacitive instruments scan the zone near the surface non-destructively and are suited to walking an area quickly; resistance measurements with electrodes give values from a defined depth but require small pin holes. Both give comparative values, not absolute ones – salts from earlier damage raise the reading of the resistance measurement, while metal and varying bulk density interfere with the capacitive one. An absolute moisture content can only be obtained by way of an oven-dry test or a carbide (CM) moisture test.

Damp building elements are cooler than dry ones because of evaporation, which is why the saturated area often stands out in the thermal image with recognisable edges. A thermal image shows temperature, not moisture – thermal bridges and heating circuits produce similar pictures, which is why every suspicion is confirmed with a moisture measurement.

If the suspicion lies in a void – suspended ceiling, dry lining, screed insulation layer, services duct – a small camera is introduced through a hole a few millimetres across. In this way standing water, run marks and corrosion become visible without opening up the construction.

If a wastewater or rainwater pipe is under suspicion, a camera is run through it: cracks, open joints, root ingress and misconnections become visible, and the defect can be marked at the surface by means of a sonde. Fine cracks in sections that are full remain hidden.

In order to rule out a pipe as the cause, the section is isolated and it is observed whether the pressure holds. This separates water from the pipework from moisture out of the structure, but it does not say where the section would be leaking.

A rigid camera head drives past the lateral connection; a multi-turn steerable head turns in and carries on – over several branches in succession as well. This makes accessible that part of the network which appears only as a dark hole in a normal survey run. Work is carried out from DN 100 to DN 200; where the head cannot get through, the recording ends and this is noted as an end point.

During the run, the distance travelled is recorded together with changes of direction and gradient, and from this the route of the network is represented digitally. The representation is a route model, not a survey drawing – for positionally accurate points, sonde location and surveying in by GPS or total station are added.

The sonde in the camera head makes the current position detectable at the surface, so that end points and notable features can be marked; for the route as a whole, electromagnetic pipe locating is added. Both react sensitively to reinforcement and services lying in parallel, which is why bearings are cross-checked.

If a section remains impassable – collapsed, obstructed, cut off – the visual recording ends. For such sections, location methods without access to the pipe come into consideration.

What passes by as a dark opening in the pipe wall during a normal survey is, in a works network, often a side run of its own: the drainage of a hall, a yard drain, a decommissioned connection. With a rigid camera head it stays exactly that – an opening in the picture. The multi-turn steerable head turns in and drives on, even across several branches in a row – turning into branches is possible from DN 100 to DN 200, driving from DN 30 to DN 300. If the head gets no further, the point reached is recorded; the section beyond stays in the report as not driven.

The camera records how far it has driven and with which changes of direction and gradient; from this the route of the network is digitally rebuilt – sections, outlets and side runs in their order and position to each other. For records management the limitation matters: this is a route model, not a survey plan. Coordinates only arise where points are pinpointed by sonde and surveyed with GPS.

Via the sonde in the camera head, the spot where the camera currently stands can be pinpointed from above and marked on the site; so every end point and every conspicuous finding gets a place that can be found again. The route of a whole run is additionally located electromagnetically. On an industrial site, reinforcement and third-party lines lie close together and distort the signal, so every pinpoint is cross-checked before it goes into the documents.

A collapsed, built-over or previously severed section cannot be recorded optically – the drive ends before it, and the gap stays in the report as a gap. Whether it can be closed depends on whether a locating method without pipe access comes into question for this section.

The crawler travels the pipe run; the movable camera head swivels and rotates so that joints, connections and lateral connections can be looked at specifically. We work from DN 100 to DN 1200, with an insertion length of up to 305 metres per access – which is part of what decides how many manholes have to be opened for a stretch.

From DN 300 the automatic lift comes into play. It raises the camera head into the pipe axis up to DN 900 so that the whole circumference is in the picture and not merely the invert – in a large cross section a camera on the floor otherwise looks past half the pipe wall. Which head travels with it depends on the nominal diameter; the large one additionally carries a tenfold mechanical zoom with which a lateral connection or a defect can be brought closer without the crawler having to travel there.

Where water submerges the camera head, the visual assessment ends.

The crawler contains a sonde whose field is picked up at the surface. In this way a distance in metres in the video becomes a point on the site – the defect is located as soon as it is seen, and not only in a second operation. Reinforcement, metal fittings and services running in parallel can distort the field; a check bearing is then taken.

On industrial and commercial sites the crawler alone is rarely enough: there, connections and branch runs hang off the collector sewers that it cannot travel. It is therefore combined with push-rod camera systems – the crawler for the pipe runs, the push-rod camera for everything that leads off them. In this way even complex pipe networks can be travelled and traced continuously instead of being broken up into unconnected individual recordings.

Coding and grading are carried out to DWA-M 149 with ISYBAU grading or to ATV-M – which set of rules applies depends on your requirement and is settled before the survey run. The findings are therefore not one operator’s estimate but a coded statement that can be compared with a later survey of the same pipe run.

If the pipe route is additionally recorded by GPS, a surveyed plan exists at the end. Without that order a plan sketch is produced that is positioned from the orthophoto – usable on site, but without coordinates.

In older existing property it is not even settled how many lateral connections there are at all – so not only the main pipe run is surveyed but every lateral the head finds: it turns in and carries on, over several branches in succession as well. What a rigid head passes as a dark hole becomes part of the recording in this way, and with it part of the drawing that did not exist until now. Work is carried out from DN 100 to DN 200. Where the head cannot get through – a blockage, ponded water, a collapsed section – the run ends; the point reached is marked at the surface and noted as an end point.

The run records the distance travelled, direction and gradient; from this the route of the network is produced as a digital representation. For a property without a drawing this is the first connected picture of any kind: which pipe run hangs on which, where the fall runs, where a lateral ends blind. It remains a route model and not a survey drawing – coordinates arise only where surveying in is added.

The recorded route is marked on tarmac, paving or grass, with the centreline, the lateral connections and depth readings – the result that counts on site. Its limit is durability: paint does not survive rain and traffic for long. That is why the route is additionally transferred into a document.

So that the route outlasts the paint, it is transferred into a document. In the open this is done by taking it off against the orthophoto: the marked points are taken off against the aerial image and the pipe run is drawn into it – building edges, paths and kerbs provide the reference. In interiors the as-built plan takes that place, and there the route is entered into the floor plan. Both are a transfer by drawing and not a survey: they show reliably where the pipe lies and where it leads, without asserting coordinates.

Where it is needed positionally correct – because there is to be building over the route, or because the route is to go into an asset record – we additionally survey the marked points in with GPS or a total station. That is a separate item and is carried out only to order; without it, the route stays as drawn in.

Hard, compact blockages and root material can be released with a rotating tool that is guided through the pipe. That is the way forward when flushing water can no longer drain away or the blockage is firmly wedged.

High-pressure cleaning is usually a method for underground drains and sewers – for the small cross-sections inside the building we use a special high-pressure unit built precisely for that purpose: with it, drains and sanitary pipes can be cleaned at high pressure. The water jet releases grease, soap and limescale deposits from the pipe wall and carries them away instead of merely boring a hole through the blockage. Afterwards the cross-section is clear, not merely passable – and that is exactly what the camera then checks.

During and after cleaning we check the result with the inspection camera. This provides evidence that the cross-section is clear, and it becomes visible whether a cause remains behind: displacement at a joint, roots through a leaking connection, back fall, broken sections.

A fog generator fills the sealed-off section with visible smoke. The smoke takes every opening that is not closed and becomes visible exactly there: at the undocumented connection that appears in no plan, at the open pipe end, at the defective joint, at the crack. The method has two limits. Under a dense surface course the smoke migrates sideways and comes out next to the defect. And it only reaches what is air-filled: what lies below water level, in an inverted siphon or behind a filled water seal, stays invisible. A pass without an escape therefore says nothing about tightness.

Where the smoke yields nothing, a test gas goes into the section and is measured at the surface with the gas detector. This takes over when paving or screed lies above the pipe or when the escaping quantity is too small for visible smoke. It is paid for with time: the area is scanned lane by lane instead of being watched as a whole.

Without sealing off, smoke or gas spreads through the whole network, and an escape no longer says which line delivers it – exactly the assignment this is about is lost. At the start therefore stand the boundary of the test section and the point where the feed goes in; in public networks this also determines which residents are to be informed beforehand. From the same definition follows which streets, yards and plant sections must be watched during the charging, and whether the emergency services get advance notice: smoke at a facade or out of a manhole cover is taken for a fire from outside.

The two methods answer different questions. The camera drives the line and shows what is going on at the pipe – whether anything leaves to the outside at that spot stays open. The smoke shows the way out, but says nothing about the kind and condition of the spot. Only both together carry the decision whether to reconnect or to repair.

A smoke generator charges the sealed-off pipe section with visible smoke, which follows every unsealed opening and becomes visible there: at pipe ends, misconnections, leaking connections and cracks. Under a sealed surface the smoke can migrate sideways, and it reaches only air-filled sections – it does not show defects below the water level, in inverted siphons or behind filled traps. A job without any escape is therefore not proof of tightness.

Instead of smoke, a test gas is introduced and detected at the surface with a gas detector. The method works where smoke would not become visible: under screed, under paving or where the quantity escaping is small; it takes more time, because the area is searched systematically.

Without controlled sealing off, smoke or gas spreads through the entire network and the findings become useless. The test section is therefore limited and a point of introduction chosen; in public networks this at the same time determines which adjoining occupiers have to be informed in advance.

The camera shows the defect from the inside, but does not say whether anything escapes there; the smoke shows the escape, but not the type and condition of the defect. Together they give the complete findings.

VLF stands for very low frequency: testing is carried out with an alternating voltage of very low frequency instead of at mains frequency. There is a design-related reason why this detour is needed at all, and it explains the choice of method.

Polymer-insulated cables must not be tested with direct voltage – direct voltage deposits space charges in the insulation and can damage the cable instead of assessing it. Testing at mains frequency would be the obvious alternative, but for a longer section it would require a reactive power that no transportable set can provide. The very low frequency solves both: the insulation is stressed in a way similar to mains operation, at a power that can be generated on site.

The test voltage and test duration follow your specification. If you do not specify any, we test to the standard.

The limitation lies in the statement itself – the test shows whether the section withstands the test voltage, not how far ageing has progressed. A section that has passed may be only marginally above the limit.

The outer plastic sheath is tested against earth, that is, the question of whether the cable is still sealed. This is the test that comes first after construction work near the route, because damage to the sheath is the most common unnoticed construction damage. The limitation: it tells you that a sheath fault is present, not where – sheath fault location follows for that.

Beyond the yes-or-no statement there are methods that describe the state of ageing rather than only the ability to withstand: partial discharge (PD) measurement, which shows weak points in the insulation before they become faults, and tan delta (dissipation factor) measurement, which uses the dissipation factor to assess ageing and moisture uptake over the entire length. Both are more involved and more demanding to evaluate than a withstand voltage test.

Coaxial cables have a defined characteristic impedance, and time domain reflectometry depends on it: every deviation – a crushed section, a damp plug connection, a torn screen – reflects part of the pulse, and the shape of the echo reveals the nature of the discontinuity. The limitation lies with long, heavily attenuated runs and with several discontinuities one after another.

A control cable offers the most favourable starting point for bridge measurement: all the cores have the same cross-section and the same length, so the reference for comparison lies alongside within the same sheath. Comparing the faulty core with a sound one gives the ratio of the lengths up to the fault position. Both ends and a sufficiently low-resistance fault are needed; an undocumented junction box falsifies the length.

Telecommunications and data cables are twisted in pairs so that interference acts equally on both cores. This lay can be put to use: a signal fed in produces a pattern at the surface in the rhythm of the twist, and this pattern ends at the fault position. With an irregular lay length or with screened designs the method does not work.

For a distance to become a point that can be marked, the route has to be known: a locating signal is applied and the route is traced together with a depth reading. Control cables of the same design cannot be told apart from the outside; cable identification decides which cable may be cut on the basis of the signal, not of its appearance.

Devices that react sensitively to voltage are connected to control and signal runs, so the initial findings determine the method. The choice of measuring method is agreed with the person responsible for your installation.

The first reading does not say where the fault is, but what one is dealing with. The insulation resistance of every core is measured against earth and against its neighbours, together with the conductor resistance. That is the point at which everything else is decided – and at the same time the documented condition of the feeder at the moment of failure, which you can take into your own operating records.

A pulse travels through the cable, every discontinuity reflects it differently, and the instrument calculates the distance from the transit time. On a supply cable the trace is usually easy to read, because there are fewer joints between two cabinets than on a medium-voltage ring that has grown over decades. What it gives is a length along the cable – not yet a position on the ground.

Bridge measurement needs no echo: it compares resistances instead of reflections and gives a distance even where reflectometry yields nothing. Where there is an earth fault on one core, the ratio of the lengths follows from the comparison with a sound core in the same cable. What is required is a reachable far end, a consistent cable construction throughout, and a fault of sufficiently low resistance.

Where the TDR trace remains ambiguous, the Ley effect helps to narrow things down. What is evaluated is how the applied signal behaves along the route: consistent up to the fault position, no longer beyond it. That is a narrowing down, not a pinpointing – it says on which section the ground is to be opened.

A sheath fault is damage to the outer covering through which moisture reaches the screen. Location works via the voltage gradient: a pulsed DC voltage drives current through the damaged point into the soil, two earth spikes pick up the gradient at the surface, and directly above the fault it reverses. On sealed surfaces and over reinforcement the gradient is blurred.

If the insulation value stays too high, the fault position yields nothing to measurement; it is then subjected to surge voltage and burned down until it becomes earth-sensitive. On low voltage this is not the normal case, but it does occur – and when it does, the feeder is completely out of service, because the neighbouring phases are frequently damaged as well. What that means for your supply, and what has to be in place beforehand, is set out in full on the page for medium voltage, where this type of fault predominates.

In the open trench several systems lie side by side, some of them live. Identification establishes which cable in the bundle is the one being sought before anybody cuts; if all the cables are of identical construction, your operations team has to be involved.

The first reading does not answer where the fault is, but what one is dealing with. Core by core, the insulation is recorded against earth and against the remaining cores, together with the resistance. The result sets the course for the entire job: a low-resistance fault can be followed up immediately, a high-resistance one cannot. For your operating records the same finding is at the same time the survey of the feeder’s condition at the moment of failure.

High-resistance faults show no echo in the TDR trace, and it is precisely this type of fault that predominates on medium-voltage cables. If the insulation value stays too high, the fault position is invisible to every measuring method. It is therefore subjected to surge voltage and burned down until the transition resistance falls far enough for it to become earth-sensitive – only then does fault location work at all. This is the step with consequences for your operation: the feeder is completely dead while it happens, and the neighbouring phases are frequently damaged as well. It therefore has to be settled beforehand how the loss of supply will be bridged – by a switchover or temporary arrangement, or by having the excavator standing ready to open up immediately after the location work.

A sheath fault is damage to the outer plastic sheath through which moisture reaches the cable screen; the installation often keeps running while corrosion is already at work. For location, a pulsed DC voltage is applied to the screen, the current leaves the cable at the damaged point and creates a voltage gradient which is picked up with two earth spikes and reverses direction above the fault position. On sealed surfaces, over reinforcement, or where the screen is earthed at several points, the gradient is blurred.

On distribution runs that have grown over time, the TDR trace often gives no unambiguous picture, because every joint and every branch split sets a marker of its own. The Ley effect does not work with echoes but with the behaviour of the applied signal along the route: consistent up to the fault position, no longer beyond it. That allows the section in which the ground is to be opened to be narrowed down – a narrowing which, together with the measured distance and the route, becomes the excavation instruction.

In the distribution network, time domain reflectometry also serves as a survey of the existing assets: the trace shows joints, branch splits and terminations as recurring markers. If the current trace is compared with an earlier measurement of the same run, the new discontinuity stands out.

Unlike time domain reflectometry, bridge measurement does not depend on an echo: it evaluates resistances instead of reflections and gives a distance even where the trace shows nothing but joints and branches. With an earth fault on one core, the ratio of the lengths follows from the comparison with a sound core of the same cable. What is needed is access to the far end, a consistent cable construction and a sufficiently low-resistance fault.

In distribution networks several systems lie side by side, some of them live. Cable identification identifies the cable being sought within the bundle beyond doubt before it is cut; if all the cables are of identical construction, the involvement of your operations team is required.

Before any location work comes the question of which fault is present: the resistance of the heating conductor is measured against the design value for the circuit, and the insulation against the protective earth braid. If the heating conductor has an open circuit, the work is based on time domain reflectometry; if there is a fault to the braid, it is based on sheath fault location. Where several circuits lie under one area, the same finding also establishes which circuit is affected at all.

The pulse travels as far as the open circuit and is reflected there; the transit time gives the length of the sound section, and in relation to the total installed length it shows which loop the fault sits in. The velocity factor (VoP) for heating conductors is not stored in the instrument and is calibrated against a sound circuit of identical construction; without this the reading in metres can only be read as a relative value, and very long heating circuits reach the limit of the measuring range. A figure only becomes a position in the room once the run from the thermal image is added to it.

The length does not only count for the location work. Heating mats come prefabricated from the manufacturer, with a defined output and a maximum load that may sit on the circuit. If the mat is shortened, the ratio of conductor length to applied output shifts – which is why the length actually installed has to be known before anybody intervenes.

If there is no break but a fault to the protective earth braid, the covering of the heating conductor is damaged. For location, a pulsed direct current is fed in which leaves the conductor at the damaged point; the voltage gradient this produces is picked up at the surface and reverses direction directly above the fault. Heavily reinforced concrete blurs this gradient – there the combination of heating up and the thermal image carries the work further.

The heating conductor lies under tiles, screed or concrete – it cannot be seen, but it can be heated. The burn-down generator applies enough power for the conductor to stand out from the surrounding construction and show through at the surface as a heat pattern. Two things become visible with it: the installed run with its loops and turning points, and the fault position itself, once energy is applied there deliberately and the spot becomes hotter than its surroundings. How quickly and how clearly this comes through depends on the depth of the construction and on the covering: a thin bed under tiles shows early, thick screed needs time.

The thermal imaging camera is the eye for the burn-down generator: it makes visible what the generator has put into the floor. The heated conductor shows as a pattern, and where the pattern breaks off, reverses or shows a locally limited excess temperature, that is where the fault sits. Without heat applied it stays silent, because a mat with an open circuit is equally cold everywhere. The two methods therefore belong together and are not used separately.

At the start stands not the where but the what: the resistance of the heating conductor is held against the circuit’s nominal value, plus the insulation against the protective braid. A severed conductor leads to time domain reflectometry, a short against the braid to sheath fault location. On an area with several circuits this finding also decides which circuit is pursued at all.

A pulse is applied to the heating conductor and returns from the break. The travel time yields the length of the still intact piece, and this length, set against the total length laid, says in which loop the fault sits. For heating conductors no velocity factor is held in the instrument; calibration is done on an identical intact circuit, otherwise the metre reading can only be read relatively. Extensive ramps and long open-area circuits push the method to its range limit – one reason to determine the affected circuit cleanly first.

If the heating conductor shorts against the protective braid instead of being broken, its sheath is damaged. A pulsed direct current exits at the damaged spot, the voltage gradient is scanned on top of the surface, and exactly above the fault it reverses direction. Ramps, however, are usually heavily reinforced, and reinforcement smears this gradient – where it does not narrow down enough, the combination of heating up and thermal imaging takes over.

Under concrete, screed or paving bed the heating conductor cannot be seen – but it can be heated. The burn-down generator applies enough power for the conductor to stand out from the surrounding build-up and show through at the top. This makes two things visible: the laid route with loops and turning points, and, with energy applied deliberately, the fault as a locally deviating spot. Open areas need more patience for this than a thin bed in a bathroom: ramp build-ups are thick, and frost or wet on the surface draws heat away.

The thermal imaging camera reads what the burn-down generator has put into the area: the heated conductor shows as a pattern, and where it breaks off or shows a tightly bounded overtemperature, the spot sits. Outdoors the image additionally depends on the state of the surface – a closed layer of snow or ice falsifies it, the section must be cleared for it. Wet ramps cool unevenly, which lengthens the recording time.

At the start the question is not where the fault lies but what the cable still has to offer. Core by core the insulation value is measured against earth and against the remaining cores, and the resistance is recorded with it. From this we read off which cores are damaged and which type of fault is present. This finding decides everything that follows – it says whether measurement can go ahead directly or whether the fault position first has to be made earth-sensitive. Without it, any location work would be a matter of luck.

Time domain reflectometry sends a short pulse into the cable and measures when the echo returns; every discontinuity reflects it differently, and from this the instrument calculates the distance. Besides the fault, the TDR trace also shows branches, cable joints and the cable end. The limitation: it gives a distance along the cable, not a position in the garden, and with high-resistance faults often no echo at all.

Bridge measurement is a comparison of resistances: at the far end the faulty core and a sound core are joined into a loop; the fault position divides it into two sections whose resistances are in the same ratio as the corresponding lengths. It is the method of choice for earth faults and core-to-core faults, particularly where time domain reflectometry yields nothing. Its limitation: it requires access to both ends, a sound return conductor of the same cross-section and a low-resistance fault.

A high-resistance fault is silent to measurement. The position is therefore subjected to surge voltage and burned down until it is low-resistance – earth-sensitive – because only then can it be located at all. This is not a minor intermediate step but an intervention in the cable: the installation is as a rule completely out of service while it happens, because the neighbouring phases are frequently damaged as well. It therefore has to be settled before the appointment whether a standby supply is available – or whether, instead of burning down, the ground is opened up and the cable repaired directly.

Where no clean reflection pattern comes about, the Ley effect helps to narrow things down: what is evaluated is how the applied locating signal changes along the route. Up to the fault position it behaves consistently and beyond it no longer does; the point at which it changes lies above the fault. That is a narrowing down and not a determination of the position to the centimetre; together with the distance known from the measurement it becomes the place at which the ground is opened.

A sheath fault is damage to the outer plastic covering through which moisture enters; the installation often continues to run while the damage progresses. For location, a pulsed DC voltage is applied; the current leaves the cable at the damaged point and creates a voltage gradient which is picked up with two earth spikes and reverses direction above the fault. Its limitation: on asphalt and concrete the gradient cannot be picked up cleanly.

A locating signal is applied to the isolated and de-energised cable and traced at the surface. This produces the route line on which the measured distance is set out, together with a depth reading. The limitation: in densely occupied trenches the signal couples across to other services and the depth reading remains an approximate value, which is why the route is cross-checked several times.

In front of the house, power, telecoms and cable TV frequently share a trench, and the cables belong to different owners. Cable identification establishes, before anything is opened up, which cable is the one being sought, so that nobody ends up holding a third party’s live cable. Its limitation: if several cables are of identical construction and in service, the network operator has to be involved.

Before any narrowing down comes the question of what one is dealing with. On the isolated section insulation and resistance are recorded core by core; from this follow the type of fault and how high-resistance it is. In a ring that has grown over time this finding has a second use: it shows whether only one position has failed or whether the cable returns poor values over its whole length – in which case fault location is only the beginning and the real question is the age of the cable.

Multi-core cables are twisted, with the cores running in a helix around the axis of the cable. If a locating signal is fed in such a way that it travels out and back along these cores, a rise and fall in the rhythm of the twist appears at the surface, and it breaks off at the fault position. The method requires a low-resistance fault between two cores of the same cable; with an earth fault or a plain open circuit the pattern does not appear, and there we work with bridge measurement or time domain reflectometry.

The Ley effect narrows down faults on multi-core buried cables when there is no clean reflection pattern: what is evaluated is the change in the locating signal along the route, which behaves characteristically as far as the fault position and no longer does so beyond it. It narrows the position down; it is not pinpointing.

Bridge measurement works with resistances, and in a ring that has grown over time this is where it is sensitive: aluminium and copper follow one another in the same run, and an unknown change of material shifts the result without this being apparent from the measurement. It is therefore only carried out after the section-by-section narrowing down, once the material and construction of the short length are known.

Time domain reflectometry gives a distance most quickly when there is a clear open circuit. Its limitation shows with a large number of branches: every outgoing connection at a column and every cable joint produces a reflection of its own and the trace becomes ambiguous – which is why it is only applied after the ring has been disconnected.

Cables do not always follow the edge of the pavement shown on the plan, so the actual route between the columns is recorded with the locating signal. At the base of the column the incoming and outgoing cables come together, often with third-party services alongside them; cable identification establishes, before anything is disconnected, which run leads to the next column.

Coaxial cables have a defined characteristic impedance, and time domain reflectometry depends on it: a short pulse is fed in, every deviation reflects part of it, and from the transit time of the echo the instrument calculates the distance. A crushed section, a damp plug connection or a torn screen each show up differently. The limitation: the method gives a distance along the cable, not a position in the garden, and with long, heavily attenuated runs or several discontinuities one after another the picture becomes unclear.

Bridge measurement is a comparison of resistances: at the far end the faulty core and a sound core of the same cable are joined into a loop; the fault position divides it into two sections whose resistances are in the same ratio as the corresponding lengths. Twisted-pair cables offer the most favourable starting point for this, because all the cores have the same cross-section and the same length, so the reference for comparison lies alongside within the same sheath. Its limitation: it requires access to both ends and a sound return conductor, and an undocumented junction box falsifies the length.

Telecommunications and data cables are twisted in pairs so that interference acts equally on both cores. This lay can be put to use: a signal fed in produces a pattern at the surface in the rhythm of the twist, and this pattern ends at the fault position. With an irregular lay length or with screened designs the method does not work.

A locating signal is applied to the isolated and de-energised line and traced at the surface. This produces the route line on which the measured distance is set out, together with a depth reading. The limitation: in densely occupied trenches the signal couples across to other services and the depth reading remains an approximate value, which is why the route is cross-checked several times.

In front of the house, power, telecoms and cable TV frequently share a trench, and the cables belong to different owners. Before anything is opened up, cable identification establishes from the signal and not from the appearance which cable is the one being sought, so that nobody ends up holding a third party’s cable or a live one. Its limitation: if several cables are of identical construction and in service, the network operator has to be involved.

Before anything can be surveyed in, it has to be found: services that will carry a signal – cables, metallic pipes, services with a tracer wire – are followed from the surface by electromagnetic locating. Locatable sondes go into plastic pipes, and whatever carries no signal at all is read by ground penetrating radar from the differences in density of the ground. That also makes recordable the assets of which neither a plan nor a recollection has survived.

A GNSS receiver on the survey pole determines the position of every located point from satellite signals, corrected via a reference correction service. In open ground this is the fastest method. Its limit is the clear sky: between the walls of a building, under dense tree canopies, in narrow street canyons or under bridges the accuracy breaks down.

Where GNSS fails, the total station measures direction and distance from a known station. It is independent of satellites and also works inside buildings and beneath overhead structures. It requires a clear line of sight to the target and control points through which the survey is tied to the coordinate reference system.

A line on its own is not yet a utility asset record. Every object carries the medium, the material, the dimension and, above all, the origin of the information: surveyed in, located or taken from an old plan. That distinction later decides how far the line on the plan can be trusted.

The handover is made as .gpkg or .shp for the GIS, as .dxf for CAD and as .pdf for everyone without system access. The formats are not equivalent: .shp limits attribute names to ten characters, .dxf carries the drawing and its layers but no attribute data – which format is the leading one we establish with your target system.

The quickest way to the manhole rarely runs across the area, but through the service. If there is an inspection opening, a cleaning pipe or an open connection at the house, we push the camera in there and drive outwards – with the fall, in the direction in which the waste water runs. The manhole is where the service arrives: at the junction, at the change of direction, where two runs come together. The image also says what is waiting there before anybody opens up. An opening the camera can pass through is needed; bends, displaced joints, roots and the reach of the rods set the limit.

The camera image on its own gives only the distance in metres, not the place in the garden. A transmitter sonde therefore sits in the camera head: it radiates from a single point, and its position can be taken at the surface together with a depth reading, down to 2.5 m. When the head stands in the junction, the operator stands above the manhole. This allows a plastic pipe on which no electromagnetic signal holds to be followed as far as the manhole as well.

If the service is metallic or has a tracer wire in it, it also works without a camera: an alternating current signal is applied to the pipe or the tracer wire and the route is followed at the surface. Where it ends, turns off or meets a second pipe run is usually where the manhole lies. This is the method with the greatest reach downwards: as far as 5 m. A connection point and a conductive path are needed; in densely occupied routes the signal couples across onto neighbouring services.

If nothing at all is known – no access, no service, no clue – the direction is reversed and the area itself is searched. The magnetometer measures the disturbance of the earth’s magnetic field caused by a ferromagnetic body. Cast iron covers and frames therefore show up under asphalt, gravel and soil as well, without any applied signal, and they do so quickly and across the area. The search depth is one metre at most – deeper than that the disturbance is lost in the earth’s field. The second limit is set by the surroundings: plastic and concrete covers stay invisible, and where reinforcement, building rubble or metal fences lie, the deflection is lost in the interference.

Where there is no iron, ground penetrating radar works the same area. An antenna sends short electromagnetic pulses into the ground; at every interface with different electrical properties, part of the pulse is reflected and recorded as a travel time. This makes concrete rings, the manhole structure and the void beneath it visible – precisely the cases in which the magnetometer finds nothing, and at two metres at most it reaches twice as deep. The limit is the ground: in cohesive or wet material the penetration depth falls, and under dense reinforcement the image becomes unreadable.

The radar reads density contrasts: masonry against soil, void against backfill, steel against concrete. Antenna choice sets depth and resolution – deep and coarse for foundations, shallow and fine for embedded objects. What cannot be assigned clearly is called suspected rather than sold as a finding; that honesty separates a usable report from a colourful graphic.

The radar reads the reflections inside the concrete: heating loops, reinforcement and conduits show up as separate layers in the depth slices. The craft lies in the interpretation – telling reinforcement from a heating pipe takes experience with exactly these areas, and our crews survey them year in, year out. The structure itself is never touched.

An alternating-current signal goes onto the cable – clamped on directly, inductively or via the signal clamp, depending on what the access point yields. The receiver picks up the magnetic field at the surface and traces the route. A metallic conductor must be present. In densely packed routes, as they occur on industrial sites and in local network sections, the signal couples onto the neighbouring lines – which cable in the bundle is the one in question is then only settled by identification in the open trench.

If the metallic conductor is missing, a battery-powered transmitting sonde rides along – on the pushrod camera or on rods. It transmits from one point, and this point can be determined precisely at the surface, including its depth. This gives non-metallic pipes in the section a traceable route too. The precondition is an opening through which the sonde can be inserted.

Some pipes give themselves away through sound – through the noise they generate themselves, or through one that is imposed on them. The ground microphone searches the surface for the maximum; that is where the pipe runs. This is the route when nothing couples electromagnetically. On operational areas and along streets, traffic noise sets the method a hard limit, as does a pipe material that swallows the sound.

If several cables lie side by side in the open trench, no locating plan answers the decisive question: which one is it? In cable identification a signal is applied and the single cable in the bundle is unambiguously identified before anyone cuts. The same applies on the cable tray, where assignment lists and labels rarely still hold after years.

For the areal search without a coupling point we use ground penetrating radar. It is the route to non-metallic pipes without access and without tracer wire, and the fallback where no signal can be applied.

Every located point is surveyed – with GPS in the open, with the total station where hall walls, vegetation or superstructures cut off satellite reception. This runs as standard and is not an extra order: only so does the route survive the next rain, and only so does your records management have something to connect to.

An alternating current signal is applied to the service, connected directly, inductively or via a signal clamp; the receiver measures the magnetic field at the surface and traces the route. The method needs a metallic conductor, and in densely occupied routes the signal couples across onto neighbouring services.

A battery-powered transmitter sonde is pushed into the pipe, on the push-rod camera or on the rods. It radiates from a single point, so its position can be determined exactly at the surface, together with a depth reading – this makes non-metallic pipes locatable as well. Without an opening it can pass through, the sonde cannot be inserted.

What is used is the noise a service produces itself or that is imposed on it; a ground microphone picks it up at the surface, and the maximum lies over the service. This helps where nothing can be coupled electromagnetically. Traffic noise and sound-absorbing pipe material set a hard limit to the method.

Where no point can be found at which a signal can be applied, the area-wide search with ground penetrating radar remains. It is the way to a plastic pipe with no access and no tracer wire – and at the same time the most difficult case, which we name as such beforehand.

If the detection work is to produce more than a marking, we record every measured point in addition: with GPS in the open, with the total station where satellite reception does not carry between house walls or under vegetation. From that we work the route up into a plan and hand it over in the format that designers and building contractors work with anyway – .dwg or .dxf. That is a separate item and only goes ahead if you order it.

An antenna sends short electromagnetic pulses into the ground. At every interface with different electrical properties – a pipe wall, a void, the edge of a foundation, a change of stratum – part of the pulse is reflected and recorded as a travel time. This lets ground penetrating radar find plastic pipes without tracer wire, concrete elements and voids as well. Its limit is the ground itself: in conductive, cohesive or water-saturated material the penetration depth falls; high antenna frequencies resolve finely but reach less deep.

On historic ground the same method becomes a preliminary investigation: old wall lines and forgotten services are mapped without any digging. The survey does not, however, date anything and does not replace an excavation.

A grid is laid over the area and scanned across its full extent; the individual lines are combined so that a map of the hidden installations emerges. In a focused area investigation, a critical part of the area is worked through with close depth resolution. The grid spacing decides what is found. The lines are run in two mutually perpendicular directions, because an object lying parallel to the scan line produces no hyperbolic reflection in the radar image.

The same equipment shows foundations, vaults, backfilled cellars and other remains of structures before a conversion or a demolition is planned – together with foundation dimensions and the location of cisterns, greywater, oil and fuel tanks. In industrial buildings we check thermally activated building components so that core drilling does not hit a heating loop.

Metallic services are additionally covered electromagnetically, pipes by means of an inserted sonde. How this works in detail is described on the page on pipe and utility detection.

A GNSS receiver on the survey pole determines the position of every located point from satellite signals, corrected via a reference correction service. Only this turns the colour marking into a data set for GIS and CAD. Between the walls of a building, under dense tree canopies or beneath overhead structures the accuracy breaks down – there the total station takes the place of satellite reception and measures direction and distance from a known station. Either way, everything we locate is surveyed in; that is not an additional order but part of the clearance survey.

Every ferromagnetic body distorts the earth’s magnetic field around it; the magnetic locator reads exactly this distortion. Cast-iron valve and hydrant caps are a small target but announce themselves – no signal needs to be applied, not even under asphalt, gravel and soil. The limits are set by the surroundings in the road space: plastic caps and concrete covers have nothing the instrument responds to, and between reinforcement, rubble and other metal the reading of such a small part drowns in the noise.

The antenna sends short electromagnetic pulses downwards; at every layer boundary with different electrical properties part of them returns and is recorded as travel time. This produces a sectional image of the build-up in which concrete rings, chamber bodies and the void beneath them stand out – the cases in which the magnetic instrument stays silent. How far this reaches is decided by the ground: in cohesive or wet material the pulses do not go as deep, and under dense reinforcement the image can no longer be read.

If the cap does not show itself, we take the route along the main: at the nearest accessible point an alternating-current signal is applied to the pipe or tracer wire, then the run is walked out on the surface. Line ends, bends and branches are the places where a fitting usually sits in a network – there the search tightens. This requires a coupling point and a continuously conductive path; on plastic pipes without a tracer wire the method is out, and in densely packed routes the signal jumps to the neighbouring line.

Where there is an opening, we insert a transmitting sonde, on rods or on the pushrod camera. The sonde transmits from one point; this point can be pinpointed from the surface and recorded with its depth – so even a plastic run can be followed metre by metre to the part in question. Without an accessible opening it does not work, and bends, offsets, roots and the length of the rods limit how far the sonde gets.

This method serves the manholes and entries of the job, not the fitting cap: from the nearest accessible manhole the camera is driven into the section. The manhole in question shows itself from inside as an entry, connection or drop, and is surveyed on the surface via the camera sonde. It is the most unambiguous method because it shows the manhole itself. It does need an accessible neighbouring manhole and a free section – if it stands full of water, the run ends before the target.

Two sensors are placed at accessible points on the same section, for example at gate valves or hydrants; from the difference in travel time of the leak noise, the position along the line is calculated, even where plastic damps the noise at the surface. What is needed is two points at which the sensors can be coupled, and reliable data on pipe length, material and diameter.

Where no correlation is made, listening is done: from surface box to surface box, every fitting individually, with the Saalbach listening stick standing on the spindle and carrying the noise from the main upwards. The comparison is not against an absolute value but against the neighbouring points – a raised level says in which direction to carry on. This is the prelocation, not the point itself: it narrows down the section in which the ground microphone then works.

Escaping water produces a flow and friction noise at the damage that spreads through the pipe wall and the ground. The ground microphone is set down on a close grid over the route and searches for the maximum; that is where the mark is set. PVC and HDPE damp the sound strongly, low operating pressure makes the noise quieter, and machine noise masks it.

Sometimes no leak noise appears at all: the loss is too small, or the main lies in a low point where the escaping water produces no usable sound. Then the way through is the gas. The section is charged with tracer gas; it rises through the ground and emerges – depending on how sealed the surface is – directly above the leak, where the detector shows it.

The section has to be taken out of service for this. That is not a side condition but the actual effort of the method: beforehand the section is shut off, all service connections are closed, and every adjoining property owner has to have been notified. Only then can the gas be introduced.

The notification is not our business. It lies with the municipality, the water supplier or the water supply supervisor – that is, with whoever knows the connections and their owners. Without this preparatory work no measurement is made; it therefore belongs in the scheduling and not on the day of measurement.

As a rule the route and the material come from your asset records. If they are missing, we can search for the main ourselves: ground radar records layer boundaries and objects in the subsurface and so makes visible even runs that appear on no plan and take no signal. In cohesive or wet ground the penetration depth decreases, and the radar image is interpreted, not read off. This is a separate item and not part of the leak location.

Hydrogen peroxide is used as a 50 per cent concentrate, which is diluted on site to the application concentration set for the section. What goes into the main is therefore a strongly diluted solution, never the concentrate: the latter is a hazardous substance, strongly oxidising and corrosive. Hydrogen peroxide acts against bacteria, viruses and biofilms and breaks down into water and oxygen, which is why the treatment is regarded as leaving no residues. It acts where it reaches: coated pipe walls, dead legs and branches with no flow have to be flushed beforehand or included.

On request we disinfect with sodium hypochlorite or chlorine dioxide, in compliance with the respective safety and environmental requirements. The disadvantage is at the same time the reason hydrogen peroxide comes first: chlorine-based products impair taste and odour.

The treatment does not end with the contact time but with complete flushing out. That the agent has been removed is evidenced by the measurement of the residual in the flushed section; this is why the discharge is settled before the start and not improvised during the measure.

Testing is done with water, not with air: the section is filled, vented, brought up to test pressure and the pressure trace is observed. Water is barely compressible, which is why even a small leak shows up as a pressure drop – the method does tell you, though, whether the section is tight, not where it is not.

The Austrian standard supplements the European specification for drinking water mains; it is what the client and the executing company refer to at acceptance. A test report without a named test basis is worth little in a dispute.

The test basis, ÖNORM EN 805 together with ÖNORM B 2538, is named in the report header with its edition; this gives the document a fixed form: section tested, material, test pressure, trace, assessment. It replaces neither the as-built records nor the hygienic release.

From the recorded trace it is assessed whether the section meets the requirement. Traces that point to remaining air, temperature effects or an isolation point that moves are equally conspicuous; in that case the cause is named and the test is repeated, instead of releasing a borderline result.

What has settled is removed first: sand, silt and coarse contamination on the compartment floor as well as deposits on walls and internal fittings. This step decides the outcome of the whole cleaning, because agents do not reach the substrate through a coated surface – it removes deposits, not germs.

At the customer’s request, a chemical cleaning with food-safe agents follows; it loosens adherent deposits and is then flushed out completely. Where the coating is damaged, the finding is documented instead of working the surface further.

After cleaning, the compartment is disinfected to remove microbiological loads. What is used is Mirasan®, made by Schermann GmbH – that is, by the firm with which we carry out reservoir hygiene together. The agent comes in several variants; which one is used is decided by the kind of soiling that shows itself in the drained compartment, not by a standard recipe.

Disinfection acts on the cleaned surface, not through deposits – which is why the order of the steps is not negotiable.

Before the return to service the compartment is checked microbiologically. The finding applies to the moment of sampling; everything after that is a matter for network operations.

Every gate valve is moved through its full travel, from open to closed and back. What is assessed is whether it is free-moving and how high the turning resistance is. A gate valve that cannot be moved is recorded as a defect instead of being forced free – the condition is assessed, not repaired.

With the gate valve closed, we listen for whether water is still flowing through: a flow noise means that the fitting does turn but does not close tight and will not isolate a section in an incident. The finding says tight or not tight, not how badly the seat is damaged.

If a surface box is missing in the field, the route of the main is recorded electromagnetically and the box is searched for along this line – this way, fittings that were covered over during road works can be found again. Without a conductive main or tracer wire, only the search via existing records and ground penetrating radar remains.

At the customer’s request, the surface boxes found and prominent points on the mains are surveyed in by satellite and processed into plans of the mains; this turns the maintenance round into a survey of the asset stock at the same time.

What is measured is not a state but a pattern: over days to weeks the consumption volumes for each zone are recorded. What counts is the minimum night flow – a segment is conspicuous if its base load does not go to zero even without draw-off. If it stays elevated, the zone is divided further by closing valves step by step (step test). The measurement says that water is being lost, not why.

A network is divided by valve positions into sections that can be observed individually – permanently as a district metered area or temporarily for the duration of the analysis. The precondition is that the gate valves are free-moving and close tight; this is where the delimitation often fails, which is why valve exercising and maintenance is often the more sensible first job.

The classification follows the water balance methodology of the rules – in Austria the ÖVGW rules on water losses – and separates billed consumption, own use, apparent losses from meter errors and real losses. Apparent losses act like real ones in terms of volume and have to be eliminated in discussion with the operator.

The profile determines the order in which segments are searched acoustically. Without this preparatory work the search through the local network is blind – every street is equally suspect, and the effort multiplies without the chance of a hit rising. With the profile the order is settled, and the pinpointing starts where the volume is missing.

A locating signal is applied to the pipe and picked up again by a receiver at the surface; from this the route and the depth follow. With metal this works directly, with plastic only by way of a tracer wire or a sonde – where several pipes lie close together, the signal couples across.

A transmitter sonde is introduced into the pipe, usually in the camera head, and located from above: a bend, a joint, a buried catchment, each in position and depth. Tight bends, silting or a closed gate valve stop the sonde.

For drinking water, separate push-rod cameras kept germ-free are used, not the equipment from the wastewater field. The image makes silting, root ingress, encrustation, displaced joints and cracks visible; a heavily silted pipe run has to be cleaned beforehand.

What happens to the located route is your decision: either matters rest with the marking at the surface, or we additionally survey the pipe in with the total station. Only the survey turns a marking that the next rain washes away into a permanent data set – it is a separate item and is carried out only to order.

Here the potable-water CCTV camera kept germ-free does not serve the search but the taking of evidence: it records how far the cross-section is reduced and where the deposits lie. If the pipe is completely blocked, the recording is made section by section from the next access point.

The cleaning is mechanical – not chemical and not by flushing. Which tool is used depends on what is sitting in the pipe: a root that has grown in through a joint or a crack holds the cross-section fast and catches new material; it has to be worked off, not pushed aside. Encrustation and settled sand call for a different tool again. Whatever cannot be released because the pipe is defective at that point is noted as a finding instead of being worked on further.

Released is not the same as carried out. For the removal, food-grade compressed air goes into the pipe: it stirs up sand, fine material and deposits, and the spring’s own water carries them away. The water that runs through the pipe in any case therefore does the transporting – which is why neither an additional flush nor water brought in is needed.

What matters is the direction: the material has to leave the installation and must not be pushed into the reservoir or the next section. Otherwise the problem merely travels on and settles again there.

The benefit shows not in the sediment released but in the volume that arrives at the bottom. A comparison that holds needs the same measuring point and the same boundary conditions: the yield varies with the season and rainfall, and a comparison across weeks is to be read only with reservations. We do not undertake to measure the flow.

After the cleaning, the same pipe run is surveyed once again. Only the pair of images from before and after makes the work verifiable – for the operator as much as for whoever releases the invoice.

A cleaning restores a cross-section, it does not replace a pipe. Cracks, displacement or leaking catchments remain documented but not remedied – the repair is carried out by your water main contractor.