A report can show enough water delivered over a week while concealing a shortfall on the day it mattered. Adding volumes first and checking demand afterwards can let a later surplus cancel an earlier deficit on paper. Whether that cancellation represents a real service depends on timing, storage and the receiving boundary.
The short answer is to calculate the relevant shortfall at the time and place where the service requirement applies, then preserve those components when summarising. A monthly total, the fraction of compliant days and the total unmet volume are different measures. Regrouping the same observations can change a reported score without changing any actual delivery.
We will work through original fictional delivery records and a simple inequality explaining the loss of information. The exercise supplies its own demand and compliance definitions. It is not an assessment of a real drinking-water service, an emergency plan or a regulatory reliability test. Water quality, access and other service conditions remain separate questions.
The Water Security subject guide follows water from source to receiving service. This branch asks a narrower reporting question: which shortfall evidence survives when fine records are turned into broader periods?
eduKateSG · Water Security
Find your next learning step
Choose the question closest to your learning need, or follow the worked examples in order. Each route connects a clearly stated model with a changed-context check.
ROUTE 1 · CHAPTERS 1–4
Define the receiving service
Name periods, units and the no-carry shortfall rule.
ROUTE 2 · CHAPTERS 5–8
Trace cancellation through reporting
Work the record, prove the grouping inequality and change boundaries.
ROUTE 3 · CHAPTERS 9–12
Choose the measure and denominator
Separate volume, block, period and duration measures.
ROUTE 4 · CHAPTERS 13–16
Preserve physical and data limits
Add an explicit storage model and preserve local and missing evidence.
ROUTE 5 · CHAPTERS 17–21
Practise and continue
Check thresholds, solve a new record and follow further reading.
Chapter contents
Define the receiving service · Chapters 1–4
1. Put the requirement at a defined receiving boundary
2. Distinguish a period amount from an instantaneous rate
Trace cancellation through reporting · Chapters 5–8
5. Group periods in pairs and locate the lost evidence
6. Prove why grouping cannot increase this netted shortfall
Choose the measure and denominator · Chapters 9–12
9. Keep volume fractions separate from period counts
10. Define what a compliant larger block means
11. Weight unequal periods by the question being asked
12. Aggregate rates through volumes rather than averaging percentages
Preserve physical and data limits · Chapters 13–16
13. Allow storage only through an explicit chronology
14. Keep backlogged demand distinct from a missed service
15. Preserve location when several receiving points are combined
Practise and continue · Chapters 17–21
17. Test thresholds and rounding before comparing reports
18. Reconstruct a changed reporting example independently
19. Diagnose the first incorrect step and preserve the source record
20. Questions about balanced totals and hidden shortfalls
21. Continue through service, storage and measurement evidence
CHAPTER 1 OF 21 · Define the receiving service
1. Put the requirement at a defined receiving boundary
For the first exercise, each period has a demand of 100 cubic metres at one receiving point. The supplied delivery value is the usable volume credited to that point during that period. No volume can be carried from one period to another under the exercise’s initial accounting rule.
This deliberately simple contract prevents ambiguity about whether a later delivery repairs an earlier service failure. It does not claim that real systems lack storage. It says storage and deferred demand must be supplied explicitly before they can change the interpretation of the record.
The WHO/UNICEF Joint Monitoring Programme includes availability when needed within its broader drinking-water service framework. Our numerical exercise does not reproduce that indicator or determine a household’s service classification. It uses the general distinction between having an aggregate amount and having a service at the relevant time.
Write down the receiving point, period and intended use before calculating. An upstream production total, a reservoir inflow and water delivered to the receiving point are not interchangeable measurements. A correct sum at the wrong boundary can leave the actual question unanswered.
For classroom learning, use the supplied record rather than inventing operational details about a real neighbourhood. The point is to learn what a summary can and cannot establish, while keeping any actual service decision attached to competent evidence and current responsible guidance.
CHAPTER 2 OF 21 · Define the receiving service
2. Distinguish a period amount from an instantaneous rate
Our first values are volumes for complete periods. They can be added because they refer to compatible amounts under one boundary. If the record instead lists cubic metres per hour, those rates must be integrated or combined with their durations before becoming volumes.
A rate of 100 cubic metres per hour maintained for two hours represents 200 cubic metres. The same rate for half an hour represents 50. Adding the two rates as 200 cubic metres per hour would not calculate the combined delivered volume.
Equal intervals make many classroom tables look simpler than the underlying question. Once interval lengths vary, an unweighted mean of rates may no longer represent the whole period. The total amount remains the sum of rate multiplied by duration when the rate is stipulated constant within each interval.
The USGS data-category explanation provides an authoritative example of why data type and statistic matter: its continuous observations and daily summary values are described separately. That documentation concerns USGS water observations, not our fictional delivery criteria, but it reinforces the need to identify what each exported value represents.
On the worksheet, put the unit in the heading and preserve the interval definition. A later reporting change from hourly rates to daily volumes should be recorded as a transformation, not a cosmetic relabelling of the same number. The service comparison depends on that distinction.
CHAPTER 3 OF 21 · Define the receiving service
3. Define shortfall and credited service before summing
Let D_t be the supplied demand volume in period t and V_t the supplied usable delivery in that same period. Under our no-carry rule, define shortfall U_t=max(D_t−V_t,0). Surplus in a period does not create a negative shortfall.
Define credited service C_t=min(D_t,V_t), assuming both inputs are nonnegative. Then C_t=D_t−U_t. These are definitions for the exercise, not assertions that every real water-performance report uses the same metric.
A period with demand 100 and delivery 80 has shortfall 20 and credited service 80. A period with demand 100 and delivery 120 has shortfall 0 and credited service 100. The extra 20 remains visible as an amount above this period’s stated demand, but it is not automatically credited to an earlier unmet need.
This clipping step matters because unmet service cannot be represented faithfully by a signed residual if the intended question is total shortfall. A negative residual from one period and a positive residual from another can cancel algebraically while describing two very different experiences.
Ask the learner to explain why max is applied before addition. If the answer is merely that the teacher’s formula says so, return to the service contract. The operation encodes the rule that an oversupply in one period does not erase a deficit in another without an explicit transfer mechanism.
CHAPTER 4 OF 21 · Define the receiving service
4. Reconstruct the four-period record
Use demand 100 in each of four equal periods. Supply deliveries 80,120,80 and 120. Total demand and total delivery are both 400 cubic metres. The equality of those totals is true and worth reporting; it is not the whole service result.
Period shortfalls are 20,0,20 and 0, totalling 40. Credited service is 80,100,80 and 100, totalling 360. Under the exercise definition, the credited-volume fraction is 360/400=90 per cent.
| Period | Demand m³ | Usable delivery m³ | Period shortfall m³ | Credited service m³ |
|---|---|---|---|---|
| 1 | 100 | 80 | 20 | 80 |
| 2 | 100 | 120 | 0 | 100 |
| 3 | 100 | 80 | 20 | 80 |
| 4 | 100 | 120 | 0 | 100 |
| Total | 400 | 400 | 40 | 360 |
Two of the four equal periods fully meet their supplied demand. The fraction of fully met periods is therefore 50 per cent. Ninety-per-cent credited volume and fifty-per-cent fully met periods are both correct descriptions under their stated definitions.
Do not average those two percentages into a single quality score. They count different things. Nor does either result prove an actual service’s adequacy, safety or compliance. The table is an accounting exercise whose value lies in keeping the measures attached to their meanings.
CHAPTER 5 OF 21 · Trace cancellation through reporting
5. Group periods in pairs and locate the lost evidence
Now combine periods 1 and 2 into Block A, and periods 3 and 4 into Block B. Each block has total demand 200 and total delivery 200. If we calculate shortfall only after grouping, each block receives max(200−200,0)=0.
That coarser calculation reports no shortfall even though the original record had 40 cubic metres of period-specific unmet demand. The later surplus in each pair has cancelled the earlier deficit within the signed sum before clipping.
| Reporting block | Demand summed first | Delivery summed first | Shortfall calculated after grouping | Original period shortfalls summed |
|---|---|---|---|---|
| A: periods 1–2 | 200 | 200 | 0 | 20 |
| B: periods 3–4 | 200 | 200 | 0 | 20 |
| Whole record | 400 | 400 | 0 | 40 |
The two shortfall columns answer different questions. One asks about net volume over each wider block. The other preserves shortfalls under the original period-specific service contract. The mistake is labelling the first as if it proved the second was zero.
A careful report can show both. It might say that aggregate volumes balanced over each two-period block, while period-specific unmet demand totalled 40 under the stipulated no-carry rule. That is more informative than hiding one valid number or declaring the other arithmetic incorrect.
CHAPTER 6 OF 21 · Trace cancellation through reporting
6. Prove why grouping cannot increase this netted shortfall
Write each signed period residual as x_t=D_t−V_t. Fine shortfall is the sum of positive parts, Σmax(x_t,0). Coarse net shortfall for one group is max(Σx_t,0).
Every x_t is less than or equal to max(x_t,0). Summing preserves that inequality. The sum of positive parts is also nonnegative. Therefore max(Σx_t,0) is less than or equal to Σmax(x_t,0).
This short proof explains the direction of information loss for our defined metric. Grouping and then clipping can hide shortfall through cancellation; it cannot make the netted positive residual larger than the sum of the original positive residuals in the same group.
Equality holds when the group’s cancellation does not reduce the positive-part total. If all residuals are nonnegative, both routes give their sum. If all are nonpositive, both give zero. A group containing both a genuine deficit and a genuine surplus produces strict loss relative to the positive-part sum, apart from zero-valued boundary cases.
The proof does not establish how an actual tank operates or whether delayed demand can be served. It is a statement about two summaries of supplied residuals. Keeping that boundary explicit lets the mathematics be exact while the engineering interpretation remains conditional on the actual service model.
CHAPTER 7 OF 21 · Trace cancellation through reporting
7. Express the hidden amount as cancellation
Within one reporting group, let P be the total positive residual and N the magnitude of the total negative residual. Both are nonnegative. The signed group residual is P−N.
The original total shortfall is P. The netted group shortfall is max(P−N,0). Their difference is P−max(P−N,0)=min(P,N). This identifies the exact amount hidden by within-group cancellation under our definitions.
In Block A, P=20 and N=20, so the hidden amount is 20. In the whole four-period record, P=40 and N=40, so 40 is hidden. The matching totals do not mean the fine record contained no period-specific deficit.
Try a group with residuals 30 and−10. Fine shortfall is 30, netted group shortfall 20 and hidden cancellation 10. With residuals 30 and−50, the coarse shortfall is 0 and hidden cancellation 30. The extra surplus beyond 30 cannot hide more shortfall than the original positive total.
This decomposition is useful for an audit because it quantifies the transformation rather than only criticising it. A report can preserve P, N and the grouping rule, allowing another reader to reconstruct both summaries. A single net residual loses the separate positive and negative components needed to recover the original shortfall.
CHAPTER 8 OF 21 · Trace cancellation through reporting
8. Change the block boundary without changing any delivery
Take the original residual sequence 20,−20,20,−20. The two blocks[1,2] and[3,4] each net to zero. Now retain period 1 as a first block, combine periods 2 and 3, and retain period 4 as a final block.
The block residuals become 20,0 and−20. Their positive parts sum to 20. The same physical record now has coarsely netted shortfall 20 instead of 0, solely because the reporting boundaries changed. The original period-specific total remains 40.
This is a useful demonstration of why report windows belong in the metadata. Two teams can compute their summaries correctly yet disagree because one uses a different partition. Without the original record or component totals, the disagreement can look like inconsistent data.
Do not choose the partition that makes performance look best and then conceal the choice. State the purpose of the reporting periods and preserve the finer service measure where it is required. A calendar convention can be reasonable while remaining different from the time scale at which a user experiences a shortfall.
For learners, draw the four periods as separate boxes before regrouping them. Write the unchanged residual inside each box. The exercise makes clear that the reporting operation changed, not the observations. It also shows why a summary cannot be compared across years if its interval definition changed without explanation.
CHAPTER 9 OF 21 · Choose the measure and denominator
9. Keep volume fractions separate from period counts
The original record has a ninety-per-cent credited-volume fraction and fifty-per-cent fully met period fraction. These measures can move independently because one weights unmet amounts while the other counts complete success at the chosen period scale.
Consider two new four-period records with demand 100 each. Record X delivers 90 in every period, giving total shortfall 40 and credited-volume fraction 90 per cent, but no fully met periods. Record Y delivers 60,100,100,100, again giving total shortfall 40 and credited fraction 90 per cent, but three of four periods fully met.
The equal volume result hides a different distribution of shortages. The different period-success result does not say which record is preferable without a purpose, severity model and receiving-service context. Our exercise supplies none of those value judgments.
Likewise, a high fraction of compliant periods can coexist with a severe failure in a high-demand period. The numerator and denominator of each measure should remain visible before any ranking is attempted.
The existing Rainwater Harvesting Mathematics guide introduces time and volumetric reliability in a tank-planning context. This branch develops a separate reporting issue: what happens to already defined shortfall evidence when its periods or aggregation operators change.
CHAPTER 10 OF 21 · Choose the measure and denominator
10. Define what a compliant larger block means
There are at least two plausible block labels, and they are not equivalent. A volume-netted compliant block has total delivery at least total demand over the block. An all-periods-compliant block has every underlying period meet its requirement.
In our original paired blocks, both blocks are volume-netted compliant. Neither is all-periods compliant, because each contains one short period. A report saying “100 per cent of blocks met demand” is therefore incomplete unless it states which rule was applied.
The original fifty-per-cent period-success result is not contradicted by either block count. The units of observation and the logical rule changed. A group-level “all” is different from a comparison of group totals, and both differ from counting the original periods individually.
Write the rule in words before coding it. For all-periods compliance, preserve the count of failing periods or an equivalent flag within each block. For volume-netted compliance, preserve the group totals and state that timing within the group is not assessed by that label.
This distinction is particularly useful when a dashboard changes from daily to weekly tiles. A greener display may reflect a broader net-volume comparison rather than improved underlying service. The correct audit question is which transformation produced the new tile, not whether green is inherently a more truthful colour.
CHAPTER 11 OF 21 · Choose the measure and denominator
11. Weight unequal periods by the question being asked
Suppose a supplied record has one one-hour period and one nine-hour period. The short period fails its stated service condition and the longer period meets it. Counting records gives one successful period out of two, or 50 per cent.
If the status is stipulated constant throughout each interval and the intended question is successful duration, the successful duration fraction is 9/(1+9)=90 per cent. This is a different denominator from a count of rows. Without the constant-status stipulation, an interval-end status would not establish every moment inside the interval.
The distinction becomes more important when a monitoring schedule changes. Splitting a long successful interval into nine one-hour records changes the row count while preserving its supplied duration. A row-count metric can then change purely because the table was represented differently.
For a time-based question, preserve durations and the relevant status evidence. For a period-based question, state the period definition. Neither approach should silently claim a finer time resolution than the observations support.
An exercise can begin with equal intervals to stabilise the idea, then deliberately remove that convenience. Ask the student which earlier fractions remain meaningful and which need redefinition. That staged change tests whether the denominator was understood or merely copied from a familiar four-row example.
CHAPTER 12 OF 21 · Choose the measure and denominator
12. Aggregate rates through volumes rather than averaging percentages
Let a first interval have demand 10 and credited service 5 cubic metres, giving 50 per cent. Let a second have demand 90 and credited service 90, giving 100 per cent. The simple average of the two percentages is 75 per cent.
The whole-record credited-volume fraction is(5+90)/(10+90)=95 per cent. It weights the interval fractions by demand volume, not equally by row. Both computations are arithmetically possible, but only the latter answers the whole-record volume question as defined.
A third report could count one fully met interval out of two, giving 50 per cent. Again the same supplied record supports different values because the targets differ. The problem is not the existence of several measures; it is presenting one label as though it uniquely identifies all of them.
When demand is zero, a per-interval credited-volume fraction would divide by zero. Do not insert 100 per cent automatically and then average it with other percentages. Define how zero-demand periods relate to the intended metric. Whole-record volume fractions can still be defined when total demand is positive.
For a reproducible summary, retain the component totals: demand, credited service and period shortfall. Rounded percentages are a poor substitute for those quantities. They can conceal both weighting choices and the exact arithmetic needed to recompute the result under an explicitly changed question.
CHAPTER 13 OF 21 · Preserve physical and data limits
13. Allow storage only through an explicit chronology
Our initial no-carry rule deliberately excluded storage. Now change the model: a suitable receiving store can hold usable surplus for a later period, subject to supplied capacity, opening stock and timing. That is a legitimate new physical mechanism, not merely a reporting preference.
In the original sequence 80,120,80,120 against 100 each, a store initially empty cannot repair the first period’s twenty-unit deficit using water that arrives in period 2. It can carry period 2’s twenty-unit surplus into period 3 under a suitable stipulated model, reducing later shortfall. The opening and closing states must be tracked.
For an explicit paper model, let current delivery serve current demand directly; opening stock covers any remaining deficit, and only the final surplus is stored up to capacity. With opening stock 20 and capacity 20, all four requirements can be met in this simplified example. The result changed because a physical state and transfer mechanism were added. Under this stipulated accounting model, opening stock 20 produces end stocks 0,20,0,20 with zero shortfall. Opening stock zero instead leaves the first shortfall 20, followed by end stocks 20,0,20. This is an aggregate teaching convention, not a guarantee that every within-period inflow schedule or outlet arrangement can realise it.
The existing sequent-peak storage guide owns cumulative storage requirements across dry sequences. Follow it when the question becomes storage sizing. This article’s main task remains preserving evidence while changing report periods.
A total-volume comparison is not a substitute for the state calculation. It does not establish that capacity, quality, access, outlet rates or opening stock support the assumed transfer. The arithmetic summary should follow the supplied physical model rather than invent a storage mechanism to justify a convenient cancellation.
CHAPTER 14 OF 21 · Preserve physical and data limits
14. Keep backlogged demand distinct from a missed service
Some fictional tasks can be deferred: a non-urgent process might accept a later delivery under a supplied rule. Other requirements are tied to the original period. These are different service contracts, and our calculations should not decide between them without evidence.
Suppose an exercise explicitly allows unmet period 1 demand to become a backlog served in period 2. A later surplus can then reduce the outstanding backlog. The report should still distinguish a delayed completion from on-time service. Both may matter to the intended user.
If the task does not allow backdating, a later volume cannot make the earlier requirement retrospectively on time. Calling the sum fully served over a wider interval answers a different question from whether the original timing obligation was met.
This distinction transfers to education and everyday planning. Completing two assignments next week does not mean both were completed by this week’s deadline. The analogy explains timing, but it does not determine which water uses are deferrable in practice. That requires the actual purpose and competent service definition.
For students, supply two versions of the same table: one with a stated backlog rule and one without. Ask for separate on-time and eventual-volume conclusions. The exercise helps them see that an assumption can legitimately change the model while a hidden assumption merely changes the story attached to unchanged numbers.
CHAPTER 15 OF 21 · Preserve physical and data limits
15. Preserve location when several receiving points are combined
The same cancellation issue can occur across places. Suppose two receiving points each require 100 cubic metres during one period. Point A receives 80 and Point B receives 120. Under a no-transfer receiving-boundary rule, total local shortfall is 20.
Combining the points first gives demand 200 and delivery 200, with net shortfall zero. That calculation does not establish that B’s extra volume was usable at A. A connection, transfer capability and timing would need to be supplied.
Do not interpret this as a claim that network transfers are impossible. It is a reminder that an aggregate boundary permits transfers on paper that a physical network may or may not support. The supply-diversity and shared-dependencies guide examines the related system-level capacity question.
For a two-dimensional teaching table, retain both period and receiving point. Calculate the defined local shortfall before aggregation when the requirement applies at that scale. A single national total can be useful while leaving local service unresolved.
The mathematical pattern is the same positive-part inequality, but the interpretation changes from time transfer to spatial transfer. Recognising that shared structure is a good transfer check. Reusing the same arithmetic does not mean reusing every physical assumption or treating a later delivery and a delivery elsewhere as identical mechanisms.
CHAPTER 16 OF 21 · Preserve physical and data limits
16. Bound what is missing instead of calling it successful
Suppose four demand periods each require 100. Deliveries are known as 80,100,unknown and 100. The unknown delivery is independently bounded between 0 and 100. Known shortfalls total 20, while the unknown period contributes between 0 and 100.
Total shortfall therefore lies between 20 and 120. Total credited service lies between 280 and 380, giving a credited-volume fraction between 70 and 95 per cent. The bounds follow from the supplied missing-value range, not from a probability model.
For the count of fully met periods, two periods are known successes, one is a known failure and one is unresolved. The successful-period fraction lies between 2/4 and 3/4. It is not automatically three out of three because one record was missing.
If the upper bound on unknown delivery were 80 instead, that period would be a known failure under the demand 100 rule even though its exact volume remained unknown. Numerical uncertainty and classification uncertainty are related but not identical questions.
Preserve the difference between no recorded delivery and a verified zero delivery. Replacing missing data with zero can overstate shortfall; excluding it without explanation can inflate apparent success. A transparent bound shows what the record supports and identifies the missing information needed to narrow the conclusion.
CHAPTER 17 OF 21 · Practise and continue
17. Test thresholds and rounding before comparing reports
Real records can include measurement uncertainty and stated tolerances. Our earlier examples used exact supplied numbers and a requirement of at least demand. A report with a different threshold must state it before classifying periods.
For an invented boundary test, let demand 100 and reported delivery 99.6. Rounding delivery to the nearest whole unit produces 100, but classification using the unrounded value still fails an exact 100 requirement. The order of rounding and comparison can therefore change the label.
Now supply an explicit teaching tolerance: delivery at least 99.5 counts as meeting the exercise criterion. The same 99.6 would pass that different rule. This is not a recommendation for a real water-service tolerance. It shows why a threshold change must not be hidden inside formatting.
If measurement evidence supports delivery only within a range 99.2 to 100.2, the exact 100 classification is unresolved under that range. Choosing the midpoint to create a definite success or failure would add an assumption. A report can retain an unresolved category and show bounds.
When comparing periods or organisations, record the threshold, unit, rounding rule and treatment of uncertain cases. A performance trend can change because of a measurement or definition change, even when the underlying service has not changed correspondingly. The purpose of the audit is to expose that possibility, not presume dishonesty whenever two percentages differ.
CHAPTER 18 OF 21 · Practise and continue
18. Reconstruct a changed reporting example independently
Try three equal periods with demand 50 each and usable deliveries 30,70 and 40. Under the no-carry rule, shortfalls are 20,0 and 10, totalling 30. Credited service totals 120 out of demand 150, giving 80 per cent. One of three periods fully meets demand.
If all periods are grouped before clipping, total delivery 140 gives net shortfall 10. The difference from the original 30 is 20. Verify it through the cancellation formula: positive residual total 30, surplus magnitude 20, hidden amount min(30,20)=20.
Now group only the first two periods and leave the third separate. The first block nets to zero shortfall and the third contributes 10, again giving 10. Group only the last two and leave the first separate: the first contributes 20 and the last pair nets to zero, giving 20. The fine-period total remains 30 in both arrangements.
For a final change, stipulate a store with opening usable volume 20 and capacity 20, under the direct-service-then-storage convention just defined. The end stocks are 0,20,10, and all supplied demands are met under that model. Track those states rather than applying the net total as a shortcut. The learner should explain which deficits the stipulated opening stock and later surplus can cover, and which final stock remains.
This set tests the operation, grouping boundary and physical model separately. A complete answer labels each result with its contract. No single percentage should be presented as the universal truth of the record while its timing rule remains unstated.
CHAPTER 19 OF 21 · Practise and continue
19. Diagnose the first incorrect step and preserve the source record
Consider the conclusion “delivery 400 equals demand 400, so every period was fully served.” The sum is correct. The first unsupported step is changing aggregate equality into a statement about all underlying periods.
A second explanation adds signed residuals first, clips at zero and calls the result total period shortfall. The operation computes a valid net-block shortfall but gives it the wrong label. Repair the order of operations or change the stated target, then preserve both measures if both are useful.
A third report compares ninety-per-cent credited volume this year with ninety-five-per-cent successful days last year. The apparent trend mixes denominators. Reconstruct comparable measures from the source record before drawing a service conclusion.
An audit card should name the receiving boundary, period lengths, demand and delivery definitions, storage or backlog rules, missing-data treatment, thresholds, component totals and reporting partition. Retain the original fine observations separately from transformed summaries and corrections.
This is the same teaching habit developed across eduKate’s first-principles articles: diagnose the exact break, retain correct work, and test the repair in a changed context. The practical goal is a report another reader can reconstruct, including what it cannot establish, rather than a reassuring percentage that has lost the evidence underneath it.
CHAPTER 20 OF 21 · Practise and continue
20. Questions about balanced totals and hidden shortfalls
Can total delivery equal total demand while some periods fail?
Yes. Our 80,120,80,120 example balances at 400 overall while period-specific shortfall totals 40 under the no-carry rule.
Is a coarser net-volume measure always wrong?
No. It answers a defined aggregate question. It becomes misleading when used as proof that every finer timing or local requirement was met.
Can storage make the later surplus useful?
Potentially, under a supplied feasible model. Opening stock, capacity, timing and receiving conditions must be tracked. Storage does not make a later delivery occur earlier.
Should shortfall percentages be averaged across reports?
Only when the target and weights justify it. Whole-record volume fractions generally need component volumes, while period or duration measures have different denominators.
What if a delivery record is missing?
Preserve the unresolved status and use justified bounds where possible. A missing observation is neither an automatic success nor a verified zero.
CHAPTER 21 OF 21 · Practise and continue
21. Continue through service, storage and measurement evidence
For the overall service chain, use Water Security and the existing supply-diversity guide. They connect source capacity, dependencies and receiving service without treating asset totals as the whole outcome.
For physical time transfer, follow Reservoirs and sequent-peak storage across dry sequences. For collection and demand modelling, the Rainwater Harvesting Mathematics guide develops its own tank context. These articles own the operating and sizing questions that this reporting lesson deliberately leaves conditional.
For authoritative context, read JMP's drinking-water service framework and USGS water-data collection categories. The former separates dimensions of service; the latter distinguishes observation and summary types. Neither source supplies or endorses our fictional delivery criteria and calculations.
Return through The Gold Standard of Education and How X Works. Carry forward the central habit: preserve the scale, boundary and order of operations that make a summary mean what its label promises.