High Water Cut Strands Central Treating Stations: Two Wellhead Short Processes Compared

Centralised treatment is the default because it works. Every well's high-water-cut mixture travels to one station, oil and water separate there, crude goes to export and produced water is polished before reinjection. The arrangement suits the medium- and low-water-cut stages well, and it keeps operating staff and maintenance resources in one place.

Then the field matures, and the same arrangement becomes the constraint.

What breaks first, and why it is not the separators

Two forces act at once. Produced liquid volume keeps growing while the water cut keeps rising, so a large share of the energy budget goes into transporting water that has no value at the station — it is pumped, heated and separated only to be pumped again for reinjection. That is a permanent loss, not an inefficiency to be trimmed.

Meanwhile the treating capacity of existing central stations approaches saturation. Some stations can be debottlenecked by reconstruction, but at a very high capital cost and with a construction window that usually has to be planned years ahead. Others have no room at all, because urban expansion has grown around them and the land simply is not available. Insufficient central-station capacity then becomes a hard limit on field profitability rather than an operating nuisance, and the pressure moves upstream: remove the water before it travels.

Pre-separation, and why the integration is the hard part

The mainstream response is to install pre-separation devices at gathering stations, or further upstream at wellheads and metering stations, so that most of the water is removed before it enters the gathering system. Existing technologies draw on gravity settling, hydrocyclone separation, flotation and thermochemical treatment, singly or in combination. Representative techniques include inclined separation, series hydrocyclones, high-efficiency three-phase separation and low-temperature demulsification.

Each has a niche, and each has a failure mode that shows up in the field rather than the brochure. Gravity settling needs residence time and floor area, both scarce at a wellhead. Single-stage hydrocyclones handle large volumes but are sensitive to flow turndown and to emulsified oil. Flotation works on the emulsified fraction but needs a downstream separation step to deal with the froth it generates. Thermochemical treatment handles difficult emulsions reliably but reintroduces chemical cost and sludge.

The unresolved gap is integration. Systematic research on combining cylindrical cyclones with hydrocyclone technology is still thin, and methods for coordinated parameter design of an integrated unit — flow split, pressure drop, retention time across both stages — still need work. Most site failures attributed to a technology are really failures of that coordination: a cyclone sized for peak flow starves the flotation stage at turndown, and the pair then performs worse than either would alone.

Two schemes built on the same core

The design logic from SINOKLE takes the patented CDFU cyclonic dissolved-air flotation oil remover as the core and adds filters of different precision to close the gap to reinjection specification. From that base, two wellhead produced-liquid short-process schemes have been put together, differing in how the polishing step is handled.

In the first scheme, the water phase leaving a two-phase separator passes through two HAC units and then two CDFU units in series. The HAC hydrocyclones take the bulk of the free water and gross oil, and the CDFU stages remove the remaining oil and suspended solids. Separated waste oil goes to a collection tank, and waste gas is routed to the flare system or the gas network. Purified water then receives advanced treatment in two stages of activated-media filters and, once in specification, is stored in the reinjection water tank. Backwash water is drawn from that tank, and the sludge produced is dewatered and shipped off site.

The second scheme keeps the same front end — two HAC units followed by the CDFU system, with waste oil and gas handled identically — but changes the tail. Effluent enters a circulation water tank, and the concentrate returns to the CDFU unit rather than being discharged; flotation scum is dewatered and shipped off site. Circulation tank effluent is then finely filtered through a SiC membrane filter to reach specification before reinjection.

The difference is worth stating plainly. The first scheme is more tolerant of variable solids load and cheaper on media replacement, at the cost of a longer train. The second recovers more water because the concentrate loops back, and its SiC membrane delivers a finer cut, but it concentrates foulants on the membrane and needs a cleaning regime the site can actually sustain.

What the two schemes buy over a conventional train

• Short process, high efficiency, strong shock resistance. Fewer unit operations between wellhead and injection well means fewer places for a slug to accumulate.
• No chemicals. Demulsification and separation are purely physical, so there is no aged oil, no secondary pollution and a cleaner environmental position. This is also what removes the dosing cost that makes conventional high-water-cut treatment expensive.
• Two-stage cyclones plus two-stage cyclonic dissolved-air flotation as pretreatment. The redundancy is deliberate: it is what holds downstream performance steady when the feed moves.
• Advanced filtration with real precision. High filtration accuracy and long media life are what convert a good average into stable compliance, which is the number a reinjection well actually cares about.
• Skid-mounted, small footprint. This is the item that decides feasibility on offshore platforms and on congested onshore pads.
• Fully automated, information-based operation with remote monitoring. Low labour requirement, and operation and maintenance that stays convenient and cheap over a field life measured in decades.

Field evidence

The configuration has been verified on produced-water reinjection duty at a PetroChina central treating station, where the skid achieved Class A1 effluent with oil content no more than 5 mg/L and suspended solids no more than 1 mg/L, and Class A2 effluent with oil content no more than 6 mg/L and suspended solids no more than 2 mg/L. Those four figures are the ones to compare against a target formation's injection specification, because A1 and A2 are not interchangeable — the tighter class costs more per cubic metre and should only be specified where the reservoir demands it.

When centralised treatment still wins

The short process is not a universal replacement. Three conditions argue for keeping the central station and accepting the transport cost. If the field is still in medium water cut, the water volume does not justify distributed equipment and the maintenance burden of many small skids exceeds the pumping saving. If well count is high and individual rates are low, the instrumentation and control cost per wellhead becomes disproportionate. And if the produced water is to be discharged rather than reinjected, the polishing duty needs a scale of plant that only a central facility can justify.

The realistic transition is staged. Install pre-separation at the wells with the highest water cut first, measure the reduction in gathering load, and let the central station recover capacity that way rather than by reconstruction. That sequence funds itself, and it keeps the central station as the polishing and backstop layer instead of forcing it to be the primary separator for water that should never have travelled.