Wednesday Reality : Further Downstream
In the summer of 1858, heat intensified the smell rising from the Thames until the occupants of the Palace of Westminster struggled to escape it. London had endured polluted water and recurring disease for years. Along the river, households and businesses lived with conditions that had become familiar through repetition. The Great Stink brought those conditions insistently into Parliament, where the prospect of continuing to work beside the river became difficult to contemplate. [1]
Below the streets, drains carried the contents of a growing city towards its watercourses. Flushing toilets could clear waste from houses more conveniently, but their contents still needed somewhere to go. Much reached a river whose tidal movement complicated its passage towards the sea. Water travelled through London carrying material that London continued to add.
The machinery of departure
Joseph Bazalgette’s intercepting sewers changed that journey. Instead of allowing local drains to discharge directly into the Thames through central London, the new network collected their flows and carried them east towards downstream outfalls. Building it required a connected understanding of the city’s terrain, drainage and underground routes. Individual drains became tributaries of a larger hydraulic arrangement. [2]
Gravity supplied much of the movement, provided the sewers maintained suitable gradients. Their depth and route therefore mattered together: every stretch had to receive what arrived from above and deliver it onwards. Where sewage needed to rise, pumping stations supplied the energy. A tunnel could look complete in isolation while still depending on machinery farther along its route before it could perform its intended function.
At Crossness, steam-driven pumps raised sewage into a covered reservoir, from which it could be released on the ebb tide. The city produced waste continuously, while the river offered a changing opportunity for discharge. The reservoir accommodated the difference between those rhythms. Its capacity allowed arrivals to continue while departures waited for the appropriate conditions. [3]
Storage, however, buys time under particular conditions. A reservoir receiving more than it can release will eventually fill unless the imbalance ends. Increasing its size extends the period it can accommodate that excess. Increasing pumping capacity changes how quickly material moves between locations. Neither establishes what the receiving water can absorb, or what happens to the material after it leaves the works.
Bazalgette’s network brought an immense public health improvement, helping separate contaminated waste from drinking water. Yet the original outfalls discharged untreated sewage, and complaints followed from communities downstream. The interception and pumping could work successfully while the receiving stretches of river remained polluted. Conveyance had accomplished something substantial. Treatment required something further. [1]
A small exception at launch
Consider an illustrative software integration connecting customer orders to a financial system. During testing, the team discovers that a particular sequence of cancellations leaves records inconsistent. The main transaction flow works. The exception affects relatively few cases. Resolving it properly would delay launch, while operations can identify the affected records and reconcile them manually.
Accepting that arrangement may make commercial sense. The launch has value, the exposure appears manageable, and a correction can follow once the service runs. Development records the limitation, operations accepts the procedure, and the programme completes its milestone. Nothing in this sequence requires negligence. The decision depends on a continuing obligation that somebody has agreed to carry.
A month later, that obligation competes with new priorities. The original developers have another commitment, while the reconciliation prevents the exception from becoming an urgent customer problem. The proposed correction waits. Operations acquires a spreadsheet, a checking routine and a small body of knowledge about which records need attention. The service remains available, and the launch continues to count as a success.
The arrangement resembles a reservoir in one important respect: spare capacity allows the system to accommodate a mismatch without immediately changing either side of it. Experienced staff provide that capacity through attention and judgement. Their intervention creates room for a temporary decision to continue. Unlike the reservoir at Crossness, however, their available capacity rarely appears on a drawing, alongside a clearly marked maximum level.
What happens during the journey
Then transaction volumes increase. Another team changes how cancellations enter the system. A new customer uses a combination of features that ordinary users seldom encounter. The reconciliation procedure needs revising, and the people who understand it begin receiving questions from colleagues who cannot safely infer what to do from the records alone.
Here the hydraulic comparison reaches a limit. An organisation can generate additional work while an unresolved issue travels through it. A rule clarified during design might have required a conversation and a change to the implementation. After launch, the same ambiguity can require incident investigation, data correction, customer correspondence, financial checks and coordination of a production release.
Delay also changes the cost of repair. Other processes may now depend on the existing behaviour. Staff have developed routines around it. Correcting the integration requires distinguishing the intended records from those already adjusted manually, then deciding what to do with the accumulated history. The original exception has acquired dependencies. Its eventual removal must account for the organisation that grew around its continued presence.
Operations consequently handles more than a fixed parcel of unfinished development. It handles the interactions between that unfinished work and a changing business. Each reconciliation can prevent harm while leaving the generating condition intact. The longer the service runs, the more evidence accumulates that the workaround functions, even as the full cost of relying on it becomes harder to reconstruct.
The department that becomes expensive
By the next budget cycle, the programme has closed. Its delivery cost remains associated with the launch, while the continuing reconciliations appear within operational expenditure. The people who approved the temporary arrangement may have moved on. Their successors inherit a functioning service and a team whose staffing requirements seem surprisingly high.
The team now needs cover for absence, documentation for unusual cases and training for new arrivals. Its dependence on experienced individuals attracts attention. Managers reasonably question why a modern integration needs so much manual supervision. The records show where the work happens, but explaining why it exists requires a journey through decisions that belong to another budget and another reporting period.
Reducing that team can remove both genuine inefficiency and the capacity compensating for weaknesses elsewhere. The distinction may become visible only after the reduction, when discrepancies reach customers or financial reporting. An incident then appears against the operational service. Unless the investigation follows the history of the obligation, the receiving department acquires another record of poor performance.
Competence can postpone that discovery. People notice unusual records, remember previous failures and intervene before an automated process produces the wrong result. Their knowledge keeps the service dependable. The more effectively they intervene, the less visible the original compromise becomes to anyone assessing the service through its outward behaviour.
What the accounts teach
Now place two programmes beside each other. One funds a durable reconciliation mechanism before launch. It takes longer and spends more within its delivery budget. The other launches with a manual procedure whose continuing expense falls elsewhere. If the comparison ends when the programmes close, the second approach can appear faster and cheaper.
That comparison can influence subsequent decisions. Future estimates face pressure from the apparent efficiency of the earlier launch. Managers who delivered within the narrower boundary receive greater credibility. The operational team asks for more capacity, but its request arrives as a recurring expense rather than part of the investment that produced the service. Separate accounts turn connected events into different management problems.
Nobody needs to falsify a number. Every recorded expense can be accurate, every milestone correctly dated, and every staffing request honestly stated. The distortion develops through the relationship between the records: costs retain their location in the organisation while losing their connection to the choices that generated them. Repetition then supplies apparently practical evidence for doing the same thing again.
Rain on the receiving ground
London’s sewers offer another complication. Combined drainage carries rainwater alongside sewage, and overflow outlets protect streets and buildings when intense rainfall exceeds capacity by allowing excess flows into the river. As more ground becomes impermeable, runoff reaches the drainage network differently. Decisions made across the city alter what the shared system must receive. [4]
The distinction matters because the receiving infrastructure encounters the combined flow. Each development can appear reasonable within its own boundaries. Together, their demands may consume capacity that none of their individual plans fully accounted for. A pipe receives water without reference to the commercial importance of the building from which it came.
Our operations team faces a comparable accumulation when several programmes each leave it a modest continuing obligation. Every exception may look manageable at approval. Their peaks can coincide with month-end reporting, staff absence or an incident elsewhere. The capacity that made the first arrangement sensible has meanwhile been committed repeatedly, without any single decision appearing large enough to explain the resulting strain.
Waiting then adds its own activity. People seek updates, investigate discrepancies twice, maintain provisional records and ask managers to negotiate priorities. The team’s workload contains both the incoming exceptions and the consequences of being unable to process them promptly. Its rising cost reflects a system changing under load, rather than a simple multiplication of the original task.
Beside the outfall
Specialisation still creates real value. A team equipped to handle exceptions can resolve them more effectively than every product team constructing its own arrangements. Launching with a temporary procedure can also preserve an opportunity that would otherwise disappear. The difficulty lies in what happens as the temporary obligation persists, demand changes and evidence of its cost travels through a different route from evidence of the launch.
At Crossness, the machinery, reservoir and tidal discharge belonged to an observable physical sequence. In an organisation, the equivalent connections can disappear behind ownership changes and reporting periods. The manual procedure becomes part of the service. The staffing requirement becomes part of the department. Eventually, even those performing the work may struggle to distinguish an unavoidable feature of the business from the residue of an earlier compromise.
Downstream, a discharge becomes part of the conditions in which people live. Those conditions acquire familiarity, along with local arrangements for managing them. Upstream, the improvements remain visible in the streets and water beside which other people conduct their lives. Beneath the ground, the network continues connecting the two, long after they have acquired separate explanations.
Historical sources
[1] Historic England — The Great Stink
[2] Institution of Civil Engineers — London sewer system
[3] Crossness Engines Trust — History of Crossness Pumping Station
Member discussion