Doc #132 — Digital-to-Print Priority Schedule

---

RECOMMENDED ACTIONS (BY ACTUAL URGENCY)

Phase 1 (Months 0–12) — The Laser Printing Window

1. [Weeks 1–4] Establish digital content extraction teams. Identify personnel with IT skills (librarians, systems administrators, database managers) and assign them to content curation — identifying, formatting, and queuing digital files for print. This is primarily a workforce allocation decision, though technical challenges exist — particularly accessing DRM-locked institutional databases, converting proprietary file formats to printable output, and establishing access credentials for institutional servers whose administrators may be unavailable. (Moderate urgency — printing hardware is functional and toner is not yet scarce, but the prioritisation framework must be in place before large-scale printing begins.)

2. [Weeks 2–8] Print the entire Recovery Library at maximum volume. The Recovery Library itself (estimated 25,000–35,000 pages across all documents) is the single highest-priority printing job.³ Target: 200–500 complete sets distributed to regional centres, hospitals, libraries, and marae across NZ. This is already covered by Doc #5 and Doc #30 — listed here because it is the first and most important item on the digital-to-print schedule.

3. [Months 1–6] Begin Priority Band 1 extraction (Section 4): medical references, pharmaceutical data, and surgical procedure guides. These are time-critical because healthcare decisions made with incorrect or missing information have immediate consequences, and the medical professionals who currently supplement their memory with digital lookups will lose that capability as devices fail.

4. [Months 3–12] Begin Priority Band 2 extraction (Section 4): agricultural guides, seed databases, veterinary references, and food preservation data. These are seasonally time-critical — agricultural decisions depend on having information available before planting seasons.

5. [Months 6–12] Conduct hardware condition survey as part of the national skills census (Doc #8). Inventory: server rooms, NAS devices, institutional hard drive arrays, and major digital collections (universities, Crown Research Institutes, LINZ, NIWA, GNS Science, public libraries). Estimate remaining operational life. Prioritise extraction from hardware showing early signs of degradation.

Phase 2 (Years 1–3) — Peak Extraction Period

6. [Years 1–2] Print Priority Bands 3–4 (Section 4): engineering data, manufacturing references, navigation tables, and maps. These become critical as NZ’s industrial self-sufficiency programme expands and maritime trade develops.

7. [Years 1–2] Prioritise internationally valuable content during the AI facility’s operational window. If the AI inference facility (Doc #129) remains operational, it could accelerate the digital-to-print programme in two ways. First, AI-assisted curation and formatting — identifying, summarising, and formatting digital content for print — could reduce the person-hours required per page of output. Second, and potentially more significant, the AI facility could produce content that does not yet exist in any digital archive: region-adapted recovery documentation for Australia, the Pacific Islands, and South America; translations of critical technical material into Spanish, Portuguese, Samoan, and other languages; and synthesised reference guides compiled from multiple digital sources. This generated content should be printed during the same window as extracted content, because it has trade and diplomatic value that could be as high as any physical export NZ puts on a vessel (Doc #129, Doc #151).

8. [Years 1–3] Transition extraction workflow to manual printing as toner stocks deplete. Digital files continue to be readable (screens, laptops) but output shifts to handwritten transcription, stencil duplication, and screen printing of key diagrams. Performance gap: Manual methods produce roughly 1–5% of laser printer throughput (Section 2.1), so only the highest-priority remaining content can be extracted during this phase. The transition also increases labour requirements per page substantially — a laser printer produces a page in seconds with minimal operator involvement, while handwritten transcription of a dense reference page may take 30–60 minutes, and stencil or screen printing requires setup time per page that only amortises at moderate run lengths (Doc #31). The binding constraint shifts from toner to person-hours.

9. [Year 2–3] Begin systematic backup of digital files to the most durable available media. Where printing is not feasible for large datasets (e.g., topographic databases, complete meteorological archives), copy files to the longest-lived storage media available — write-once optical discs (M-DISC or standard DVD-R stored cool and dry) and multiple redundant hard drives stored unpowered in controlled environments.⁴

Phase 3 (Years 3–7) — Closing Window

10. [Years 3–5] Extract remaining Priority Band 5 content (Section 4): educational curricula, technical training materials, historical and cultural archives. By this phase, toner printing is largely exhausted. Extraction means either manual transcription from screens or selective printing on domestically produced paper with domestically produced ink (Doc #29).

11. [Years 5–7] Final hardware triage. Identify the last functional digital storage devices. Concentrate remaining extraction effort on irreplaceable content — unique datasets, NZ-specific research, institutional records — that exists only in digital form.

---

ECONOMIC JUSTIFICATION

Person-years of labour

The digital-to-print programme is primarily a labour cost, not a materials cost. The materials (paper, toner) are already allocated through the printing supply strategy (Doc #5). The additional cost is the human effort to curate, format, and manage the extraction.

Content curation and formatting teams:

• Phase 1: 20–50 people full-time (librarians, IT staff, subject-matter experts reviewing content for accuracy and priority). Approximately 20–50 person-years.
• Phase 2: 30–80 people as the programme scales and manual methods require more labour per page. Approximately 60–240 person-years over 2 years.
• Phase 3: 10–30 people for final extraction. Approximately 30–120 person-years over 4 years.

Total estimated labour: 110–410 person-years over 7 years, with the majority concentrated in Phase 2.

The labour cost is front-loaded with scarce specialists: Phase 1 requires librarians, IT staff, and subject-matter experts who can curate and prioritise digital content — skills that are in limited supply and cannot be quickly trained. As the programme progresses into Phases 2–3, the work shifts increasingly toward manual transcription, binding, and distribution — general labour that can be drawn from a large available workforce.

What this produces

• Physical copies of the most important 5–15% of NZ’s accessible digital knowledge, preserved indefinitely on paper
• An estimated 10–50 million printed pages of reference material across all priority bands (the range depends on toner availability, workforce allocation, and how aggressively the programme is prioritised)
• Distributed knowledge that survives complete electronics failure — readable by anyone with eyes and light, requiring no power, no spare parts, no technical skill to access

Cost of not doing this

Without systematic digital-to-print extraction:

• NZ’s knowledge base contracts to whatever was already in print before the event, plus whatever the Recovery Library covers
• Pre-event NZ print holdings are substantial (public libraries hold approximately 8–12 million physical volumes nationally, university libraries hold millions more), but these collections are heavily skewed toward fiction, general non-fiction, and outdated editions.⁵ The technical depth NZ needs — current pharmaceutical formularies, surgical atlases, engineering handbooks, agricultural research, materials science data — exists overwhelmingly in digital form
• Medical professionals lose access to drug interaction databases, dosing calculators, and diagnostic references that they consult daily on digital devices
• Engineers lose access to materials specifications, structural calculations, and manufacturing data
• Once the hardware storing that knowledge fails, the data is unrecoverable — there is no mechanism to regenerate it from other sources

The comparison: 110–410 person-years over 7 years to preserve the most critical fraction of NZ’s digital knowledge is a modest investment relative to the cost of losing that knowledge entirely. For context, NZ’s pre-war library and archives workforce was approximately 5,000–8,000 people.⁶ Redirecting 1–5% of that workforce to digital extraction is a minor reallocation.

---

1. THE DEGRADATION TIMELINE

1.1 What NZ has stored digitally

NZ’s digital knowledge exists across several infrastructure layers, each with different failure characteristics:

Institutional servers and data centres:

• Universities (8 universities, each with terabytes of research data, digital library subscriptions, and institutional repositories)
• Crown Research Institutes (NIWA, GNS Science, Scion, AgResearch, Plant & Food Research, ESR, Manaaki Whenua — collectively holding NZ’s most important scientific datasets)
• Government agencies (LINZ topographic data, Stats NZ census data, MPI agricultural data, MBIE standards and regulations)
• District Health Boards / Te Whatu Ora (patient records, clinical guidelines, pharmaceutical databases)
• National Library / Alexander Turnbull Library (digitised collections, NZ heritage materials)
• Public library systems (catalogue data, e-book licences — the licences are worthless without internet, but locally cached content may exist)

Personal and institutional computers:

• An estimated 2–3 million laptops and desktops in NZ households, businesses, and institutions⁷
• Each contains some quantity of locally stored documents, downloaded references, cached web pages, and software
• The aggregate is enormous but fragmented, poorly organised, and mostly inaccessible to systematic extraction

Portable storage:

• USB drives, external hard drives, SD cards — widely distributed, individually small, collectively significant
• These are the most portable and most easily damaged storage media

1.2 How electronics fail

The failure modes are well understood:⁸

Component	Failure mode	Typical lifespan	NZ-specific notes
Spinning hard drives (HDD)	Mechanical: bearing wear, head crash, spindle motor failure	3–6 years continuous use; longer if stored unpowered	NZ’s humid climate (especially North Island) accelerates corrosion of internal components in uncontrolled environments
Solid-state drives (SSD)	Charge leakage in NAND cells; controller failure	5–10 years active; stored data degrades over 1–5 years unpowered depending on temperature	Higher temperatures accelerate charge loss; NZ’s temperate climate is relatively favourable
Server power supplies	Capacitor ageing (electrolytic capacitors dry out)	5–10 years	Replaceable between identical models; no domestic production
Fans (servers, desktops)	Bearing wear, dust accumulation	3–7 years	Failure causes thermal shutdown; can sometimes be replaced with scavenged units
LCD/LED screens	Backlight failure, driver board failure, capacitor ageing	5–15 years	Screen failure does not destroy the stored data — the drive can be removed and read on another machine
Laptop batteries	Capacity degradation, cell failure	2–5 years to unusable	Laptops can operate on mains power without batteries
Network switches/routers	Power supply failure, capacitor ageing	5–15 years	Needed only for networked access; standalone machines work without network

The aggregate picture: NZ’s digital storage does not fail on a single date. It decays along a curve. Assuming a mixed-age fleet of hardware at the time of the event, roughly:

• Year 0–3: 10–25% of devices fail (those already old or marginal at the event)
• Year 3–7: An additional 25–40% fail (mid-life hardware reaching end of life)
• Year 7–15: An additional 20–30% fail (the newest and most robust hardware)
• Year 15+: Scattered survivors — a few robust machines, some solid-state media, optical discs

These figures are estimates based on general reliability engineering data.⁹ The actual curve depends on the age distribution and maintenance state of NZ’s hardware fleet, which the skills census (Doc #8) would establish.

1.3 What accelerates failure

• Power fluctuations: Grid instability (voltage spikes, brownouts) damages electronics. NZ’s grid is expected to remain stable under the baseline scenario (Doc #67), but localised fluctuations are possible, especially as grid maintenance personnel are stretched thin.
• Humidity and temperature: Uncontrolled environments — warehouses, garages, unheated buildings — accelerate corrosion and capacitor degradation. NZ’s North Island humidity is a particular risk.
• Physical damage: Devices dropped, water-damaged, or cannibalised for parts.
• Use without maintenance: Servers running continuously without cleaning (dust accumulation) or fan replacement.

1.4 What slows failure

• Climate-controlled storage: Keeping critical servers and storage devices in stable, dry, cool (18–22 degrees C) environments extends life significantly.
• Power conditioning: UPS units and voltage regulators protect against grid fluctuations. Existing UPS batteries will fail within 2–5 years, but the voltage regulation circuitry remains functional on mains power.
• Reduced duty cycles: Devices powered on only when needed, rather than running continuously, extend mechanical and thermal life. This is especially important for spinning hard drives.
• Cannibalisation: Scavenging working components (power supplies, fans, drives) from failed machines to keep others running. This extends the fleet’s functional life but accelerates the rate of total loss as the donor pool shrinks.

---

2. PRINTING CAPACITY VS. TIMELINE

2.1 The printing capacity curve

NZ’s printing capacity follows its own decline curve, driven by consumable depletion:

Phase 1 (Months 0–12): Peak capacity

• NZ’s installed fleet of 30,000–60,000 commercial MFCs and 200,000–500,000 desktop laser printers (Doc #5) provides enormous throughput
• At even modest utilisation (5% of fleet active), output capacity is millions of pages per day
• Constraint: toner supply, not hardware. National toner stocks estimated at 80,000–350,000 cartridges (Doc #5)
• With print optimization (Doc #30), estimated total production capacity: 800 million to 5+ billion pages from existing stocks

Phase 2 (Years 1–3): Declining capacity

• Toner stocks depleting. Printer hardware beginning to fail (drums, fusers, pickup rollers)
• Fleet consolidation: concentrate remaining toner and functional printers at regional production centres
• Output declines from millions to tens of thousands of pages per day nationally
• Manual printing methods (stencil duplication, screen printing) beginning to supplement

Phase 3 (Years 3–7): Transition to manual

• Toner largely or fully exhausted at most sites
• Manual printing methods (Doc #29, Doc #31) producing hundreds to low thousands of pages per day nationally
• Digital-to-print extraction shifts from laser printing to manual transcription and screen printing

Phase 4+ (Years 7–15): Manual era

• Printing capacity fully domestic — paper from Kinleith mill or smaller regional pulp operations (Doc #29), ink from linseed oil (requires flax seed cultivation, oil pressing, and heat-bodying to achieve suitable viscosity) and lamp black (requires controlled incomplete combustion of resinous wood or oil — Doc #29 details the full ink production chain), printing by letterpress, screen, and stencil (Doc #31)
• Digital sources increasingly unavailable — extraction becomes a salvage operation from surviving hardware

2.2 The window

The overlap between “digital sources readable” and “printing capacity available” defines the extraction window.

Year:     0   1   2   3   4   5   6   7   8   9  10  11  12  13  14  15
Digital:  ████████████████████████████████████████████████░░░░░░░░░░░░░░
Laser:    ████████████████████████░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░░
Manual:   ░░░░░░░░░░░░████████████████████████████████████████████████████
Overlap:  ████████████████████████████████████████████████░░░░░░░░░░░░░░
                   PEAK EXTRACTION WINDOW

(█ = high capacity; ░ = low or declining capacity)

The peak extraction window is approximately Years 0–5. During this period, both digital storage and laser printing are broadly functional. After Year 5, laser printing is largely gone and manual methods cannot match the extraction rate. After Year 10–12, digital storage itself is failing at scale.

Critical implication: Content that is not printed or otherwise transferred to durable media by approximately Year 5–7 has a high probability of being permanently lost. Every month of delay in beginning systematic extraction narrows the window.

---

3. FORMAT OPTIMIZATION FOR LONGEVITY

3.1 Paper longevity

Paper is a proven archival medium. Books printed 500 years ago remain readable. Modern acid-free paper, stored in controlled conditions, has a life expectancy measured in centuries.¹⁰ Even standard office paper stored indoors lasts decades.

For maximum longevity of printed extraction output:

• Use acid-free paper where available (most modern office paper manufactured in NZ and Australia is acid-free or low-acid; verify stock with major NZ distributors such as OfficeMax/Winc and Paper Plus)
• Store printed materials in dry, cool, dark environments (away from direct sunlight, which degrades paper and fades toner) — NZ’s South Island inland locations offer naturally favourable conditions; North Island coastal sites require dehumidification
• Bind documents to prevent page loss and physical damage (Doc #30, Section 6)
• Laminate critical single-sheet references (Doc #30, Section 7)

3.2 Page density and legibility trade-offs

The digital-to-print programme faces a fundamental tension: denser pages preserve more information per sheet (conserving scarce toner and paper) but become harder to read, harder to photocopy, and more likely to require reprinting due to errors.

Recommended density by content type:

Content type	Font	Size	Columns	Margins	Est. chars/page	Notes
Prose text (medical guides, procedures)	Garamond	10.5 pt	1	Standard (Doc #30)	3,000–3,500	Readability paramount
Reference tables (drug dosing, materials data)	Garamond/Calibri	8–9 pt	1–2	Reduced	4,500–6,000	Lookup use; smaller type acceptable
Navigation/engineering tables	Calibri	8 pt	2–3	Minimum	6,000–8,000	Numeric precision required; clear column separation
Maps and diagrams	N/A	N/A	N/A	Minimum	N/A	Print at maximum resolution on best available paper
Index/catalogue pages	Garamond	8 pt	2–3	Reduced	5,000–7,000	Scanning/lookup; dense is acceptable

Two-column layout is strongly recommended for reference material. It increases content per page by an estimated 10–20% compared to single-column at the same font size (the range depends on line spacing, hyphenation frequency, and column gutter width — wider gutters reduce the gain). Two-column layout is also more readable for reference lookups than single-column, because the eye traverses shorter lines (45–75 characters per line, which is within the optimal range for reading comprehension).¹¹

3.3 Illustrations and diagrams

Medical diagrams, anatomical illustrations, engineering schematics, botanical identification images, and maps are among the highest-value content for printing — and the most toner-intensive. A single full-page anatomical diagram may consume as much toner as 5–15 pages of text, depending on tonal density and coverage area.¹²

Guidelines:

• Print diagrams at the minimum size that preserves usability. A surgical illustration that works at A5 should be printed at A5, not full A4.
• Convert colour diagrams to high-contrast greyscale where colour does not convey essential information.
• Reserve colour printing for content where colour is diagnostic: tissue identification, mineral identification, wiring colour codes, map elevation coding.
• For maps: print at the largest feasible scale on the largest available paper (A3 where possible). Maps lose critical information at reduced scale.
• For botanical/zoological identification: colour reproduction matters for distinguishing edible from toxic species. These are among the few categories where colour toner expenditure is justified.

3.4 Redundancy and distribution

A single copy of any printed document is a single point of failure. The extraction programme must produce multiple copies distributed geographically.

Minimum distribution targets:

Priority Band	Copies per document	Distribution
Band 1 (Medical)	50–200	Every hospital, major medical centre, and regional civil defence hub
Band 2 (Agricultural)	30–100	Every regional council, agricultural extension office, and major farming community
Band 3 (Engineering)	20–50	Major workshops, factories, power stations, and technical training centres
Band 4 (Navigation/Maps)	30–100	Ports, coast guard stations, regional centres, and maritime training facilities
Band 5 (Educational/Cultural)	10–30	Regional libraries and educational institutions

These targets are aspirational and depend on toner availability. The priority schedule (Section 4) is designed so that if toner runs out earlier than expected, the most critical content has already been printed.

---

4. PRIORITY BANDS

4.1 Prioritisation principles

The ordering reflects three factors:

1. Consequence of loss: What happens if this knowledge is unavailable? Medical knowledge loss causes preventable deaths. Agricultural knowledge loss causes food production failures. Historical knowledge loss is regrettable but not immediately dangerous.

2. Availability in existing print: Knowledge already widely available in NZ libraries and bookshops in physical form is lower priority for extraction than knowledge that exists only or primarily in digital form.

3. Shelf life of relevance: Knowledge needed immediately (Phase 1–2) is higher priority than knowledge needed in Phase 5–6, because the extraction window may not last that long.

4.2 Band 1 — Medical and Pharmaceutical (Print immediately)

Estimated volume: 500,000–1,500,000 pages across all items and copies

Content	Source	Est. pages (master)	Rationale
New Zealand Formulary (complete)	NZ Formulary online database	3,000–5,000	Drug dosing, interactions, contraindications — consulted daily by every prescribing clinician. Exists primarily in digital form.13
Surgical procedure atlas (major procedures)	Digital surgical references, institutional subscriptions	5,000–10,000	Illustrated guides for procedures that NZ surgeons will need to perform without specialist referral. See Doc #20.
Emergency medicine handbook	Best Practice Advocacy Centre (bpacNZ), Starship clinical guidelines	1,000–2,000	Paediatric and adult emergency protocols.
Infectious disease management	ESR surveillance data, WHO guidelines (cached)	1,000–2,000	Antibiotic selection, infection control, outbreak management.
Diagnostic imaging interpretation guide	Institutional radiology teaching files	2,000–4,000	While imaging equipment functions, interpretation skill must persist.
Dental procedure manual	NZ dental school curricula	500–1,000	Dental care without referral to specialists.
Mental health and psychiatric medication guide	NZ Formulary, Te Whatu Ora clinical guidelines	500–1,000	Psychiatric medication tapering protocols are critical as stocks deplete (Doc #122).
Veterinary formulary and procedures	MPI resources, Massey veterinary school	2,000–4,000	Livestock health management (Doc #124, Doc #122).
Midwifery and obstetric reference	NZ College of Midwives guidelines	500–1,000	Maternal and neonatal care without hospital infrastructure.

Why Band 1 is urgent: Medical professionals currently rely on digital point-of-care references for dosing, drug interactions, and procedural guidance. These are not memorised — they are looked up. When the lookup device fails, the clinician is working from memory, which is unreliable for the precision required in pharmaceutical dosing. Printing the NZ Formulary in the first months — while toner is abundant and printers are functional — is one of the highest-value uses of the laser printing window.

Existing print gap: NZ hospitals do maintain some printed references (the British National Formulary and its NZ supplement have been printed historically), but the trend over the past decade has been toward digital-only access. Many clinical guidelines, interaction databases, and procedural references exist only in digital form or in outdated print editions.¹⁴

4.3 Band 2 — Agricultural and Food Production (Print in first year)

Estimated volume: 300,000–800,000 pages

Content	Source	Est. pages (master)	Rationale
Crop production guides (all NZ-viable crops)	Plant & Food Research, FAR, MPI	2,000–4,000	Planting, cultivation, harvest, pest management for every crop NZ can grow under nuclear winter (Doc #77).
Seed identification and saving guide	Heritage seed networks, Koanga Gardens data, agronomy texts	1,000–2,000	Essential for Doc #77 (seed preservation).
Pest and disease identification (crops and livestock)	MPI biosecurity resources, AgResearch	2,000–4,000	Identification guides with images — pest management without imported chemicals.
Soil management and composting	Manaaki Whenua — Landcare Research, regional council data	1,000–2,000	Fertility management without imported fertiliser.
Food preservation methods	NZ food science literature, FAO references	500–1,000	Smoking, salting, drying, fermentation, canning procedures (Doc #78).
Fisheries species identification and management	NIWA, MPI fisheries data	1,000–2,000	Species, catch methods, sustainability guidelines (Doc #78).
Pasture management data tables	DairyNZ, Beef + Lamb NZ (Doc #75)	500–1,000	Growth rate tables, stocking rate calculators, regional guides.

4.4 Band 3 — Engineering and Manufacturing (Print Years 1–2)

Estimated volume: 400,000–1,200,000 pages

Content	Source	Est. pages (master)	Rationale
Materials properties databases	MatWeb (cached), NZ engineering standards	5,000–10,000	Tensile strength, thermal conductivity, chemical resistance — essential for machine shop operations (Doc #91) and fabrication.
Mechanical engineering handbooks	Machinery’s Handbook, institutional copies	3,000–5,000	Thread tables, tolerance charts, gear calculations, bearing specifications. Already exists in some print copies but most NZ workshops rely on digital.15
Electrical engineering reference	NZ wiring regulations (AS/NZS 3000), grid operations data (Doc #67)	2,000–4,000	Wiring standards, transformer specifications, generator maintenance.
Chemical process data	Perry’s Chemical Engineers’ Handbook, institutional copies	5,000–10,000	Process engineering data for local chemical production (Doc #17 and related).
Welding procedures and electrode specifications	NZ welding standards, fabrication references	500–1,000	Doc #94.
Plumbing and water treatment	NZ Building Code references, water treatment chemistry	1,000–2,000	Potable water systems, sewage management.
Metallurgy and foundry reference	Foundry practice references (Doc #93)	1,000–2,000	Alloy compositions, melting points, casting procedures.

4.5 Band 4 — Navigation, Maps, and Geospatial Data (Print Years 1–3)

Estimated volume: 200,000–600,000 pages (maps are page-intensive)

Content	Source	Est. pages (master)	Rationale
NZ topographic map series (1:50,000)	LINZ Topo50 series	~1,000 map sheets	NZ’s complete topographic coverage. LINZ data is digitally accessible. Printed maps are available but not universally distributed.16
NZ nautical charts	LINZ hydrographic data	200–400 charts	Coastal navigation, harbour approaches, depth soundings. Critical for maritime trade (Doc #139, #141).
Celestial navigation tables	Precomputed (Doc #10, #141)	500–1,000	Southern Hemisphere star positions, sight reduction tables, almanac data.
Tide tables (long-range computed)	LINZ, NIWA tidal models	200–500	Pre-computed tide predictions for NZ ports, extended as far forward as data allows.
NZ geological maps	GNS Science	100–200 sheets	Mineral resource locations, fault lines, geothermal areas.
Climate and weather reference data	NIWA climate database	500–1,000	Historical weather patterns by region — valuable for agricultural planning under changed conditions.

Why maps are a special case: Maps are among the most toner-intensive items to print (complex graphics, full-page coverage, often require colour) but are also among the most irreplaceable. NZ’s complete topographic coverage exists digitally at LINZ, but physical printed copies of the complete Topo50 series may not exist in sufficient numbers for post-event distribution.¹⁷ Printing even a subset — the most operationally critical sheets covering major routes, population centres, agricultural areas, and resource sites — is a high-value use of the printing window.

4.6 Band 5 — Education, Culture, and General Reference (Print Years 2–5)

Estimated volume: 500,000–2,000,000 pages

Content	Source	Est. pages (master)	Rationale
Mathematics and science textbooks	University curricula, NZ educational publishers	5,000–10,000	Secondary and tertiary level science and mathematics — the foundation for all technical recovery.
Trade training curricula	NZQA unit standards, polytechnic course materials (Doc #157)	3,000–6,000	Carpentry, plumbing, electrical, welding, machining — structured training materials.
Te reo Maori language resources	NZ educational resources, dictionaries, grammars, oral history recordings	1,000–2,000	Cultural preservation; low resource cost; appropriate only after Bands 1–4 complete. Te reo has undergone significant revitalisation since the 1980s and much supporting infrastructure is digital.18 Priority actions: print the most comprehensive available dictionary and grammar, transcribe key oral history recordings (audio/video) to written text before playback equipment fails, and print immersion school (kura kaupapa) curricula. Allocate specific curation capacity to iwi and Maori organisations — practitioners are best positioned to identify which digital resources add value beyond what oral and community knowledge already preserves.
NZ legal and constitutional documents	Parliamentary Counsel Office, NZ Legislation website	2,000–5,000	Key statutes, the Treaty of Waitangi / Te Tiriti, Civil Defence Emergency Management Act, essential regulations.
NZ history and cultural heritage	National Library, Alexander Turnbull Library digitised collections	Variable	Selective — prioritise taonga (treasured items) and unique NZ content.
Encyclopedia / general reference	Locally cached reference works	10,000–30,000	A general-purpose knowledge reference — even a partial extract of a comprehensive encyclopedia has enduring value.

Band 5 is lower priority because: Much of this content exists in NZ’s physical library collections. NZ public libraries hold approximately 8–12 million volumes, and university libraries hold millions more.¹⁹ The educational and cultural knowledge base is not as dependent on digital extraction as medical and engineering data. However, specific digital-only content (NZQA unit standards, digitised heritage collections, recent scientific publications) warrants extraction.

---

5. ESTIMATED PAGE COUNTS AND TONER ALLOCATION

5.1 Total extraction estimate

Band	Master pages	Copies	Total pages	% of toner budget
1 — Medical	15,000–30,000	50–200	750,000–6,000,000	15–25%
2 — Agricultural	8,000–16,000	30–100	240,000–1,600,000	5–10%
3 — Engineering	17,000–34,000	20–50	340,000–1,700,000	7–12%
4 — Navigation/Maps	2,000–3,100 sheets	30–100	60,000–310,000	5–15% (maps are toner-heavy)
5 — Education/Culture	21,000–53,000	10–30	210,000–1,590,000	4–8%
Recovery Library	25,000–35,000	200–500	5,000,000–17,500,000	25–40%
Operational printing (admin, forms)	Ongoing	Ongoing	Variable	10–20%
Total			~7–29 million pages	~100%

The wide range reflects uncertainty in toner stocks and printing efficiency. At the low end (80,000 cartridges, pessimistic yield), NZ produces approximately 800 million pages total — the extraction programme gets perhaps 500–700 million of those after operational printing needs. At the high end (350,000 cartridges, optimistic yield), NZ produces 5+ billion pages, and the extraction programme is limited by curation speed, not toner.

5.2 Toner allocation framework

Doc #5 establishes the national printing authority and toner rationing framework. The digital-to-print schedule integrates with this through a simple allocation rule:

Priority order for toner allocation:

1. Recovery Library production (already scheduled, Doc #5)
2. Band 1 extraction (medical — print concurrently with Recovery Library)
3. Operational government printing (rationing documents, administrative forms, public communications)
4. Band 2 extraction (agricultural — begin as Band 1 completes)
5. Band 3 extraction (engineering)
6. Band 4 extraction (navigation/maps)
7. Band 5 extraction (education/culture)

The critical decision: How much of the national toner supply to allocate to digital extraction versus other uses (administrative printing, document updates, public communications). This document recommends allocating 60–75% of the total toner budget to Recovery Library production and digital extraction combined. The remaining 25–40% covers operational needs. The exact split should be determined by the national printing authority based on actual toner stock levels established by the skills census (Doc #8).

---

6. COORDINATION WITH EXISTING PRINTING DOCUMENTS

6.1 Relationship to Doc #29 (National Printing Plan)

Doc #29 covers the physical infrastructure of printing — paper production, ink manufacture, print shop operations, and the transition from imported to domestic materials. This document (Doc #132) provides the content schedule that Doc #29’s infrastructure produces. The relationship is:

• Doc #29: How to print (infrastructure, materials, methods)
• Doc #132: What to print and when (content priority, extraction schedule)
• Together: the complete printing strategy for knowledge preservation

6.2 Relationship to Doc #30 (Print Optimization)

Doc #30 specifies the format optimization — fonts, margins, duplex settings, toner-save modes, document quality tiers — that maximise output from finite supplies. This document adopts Doc #30’s specifications directly:

• Band 1 medical references: Tier 1 print quality (full density, no toner-save — accuracy is critical)
• Bands 2–4 reference material: Tier 2 print quality (toner-save on, Garamond 10.5 pt, duplex, reduced margins)
• Band 5 general reference: Tier 2 or Tier 3 depending on expected use life
• Index and catalogue pages: Tier 3 (maximum density)

6.3 Relationship to Doc #5 (Printing Supply Requisition)

Doc #5 covers the requisition and management of existing printing supplies — toner cartridges, paper stocks, printer parts. The digital-to-print schedule is one of the demand streams drawing on the supply that Doc #5 manages. The national printing authority (established by Doc #5) is the decision-maker for toner allocation between competing demands.

---

7. STORAGE AND DISTRIBUTION OF PRINTED MATERIALS

7.1 Physical storage requirements

The extraction programme will produce an estimated 7–29 million pages of reference material (Section 5.1). At standard A4 dimensions and typical binding, this occupies approximately:

• 70,000–290,000 reams of paper (at 100 pages/ream equivalent, bound)
• Roughly 1,400–5,800 standard archive boxes
• Approximately 50–200 cubic metres of shelving space nationally (assuming standard archive shelving at ~200 pages per linear centimetre, bound in A4 volumes)

This is well within NZ’s existing library and archive storage capacity. NZ’s public library network has an estimated 200+ km of linear shelving nationally, and university libraries add substantially to this total.²⁰ The extraction programme’s storage requirements are a small fraction of available capacity.

7.2 Distribution strategy

Regional hubs: Establish at least 16 regional reference collections, co-located with the regional centres identified in the Recovery Library distribution plan:

• North Island: Auckland, Hamilton, Tauranga, Rotorua, Napier, New Plymouth, Palmerston North, Wellington
• South Island: Nelson, Christchurch, Timaru, Dunedin, Invercargill, Greymouth
• Remote: Gisborne, Whangarei

Each hub receives a complete or near-complete set of all extraction output. Within each region, subset copies (particularly Band 1 medical and Band 2 agricultural) are distributed further to hospitals, clinics, farming cooperatives, and community centres.

7.3 Catalogue and index

Every extracted document must be catalogued — title, source, date of extraction, number of copies, distribution locations. Without a catalogue, a million printed pages are useless if nobody can find the specific reference they need.

The extraction catalogue should follow the structure of the Recovery Library catalogue (catalog.md), with entries organised by subject area, cross-referenced by priority band, and indexed by keyword. Print the catalogue itself in sufficient copies for every distribution hub.

---

8. CRITICAL UNCERTAINTIES

Uncertainty	Range	Impact if wrong
Actual NZ toner stock levels	80,000–350,000 cartridges (Doc #5)	Determines total extraction volume — at the low end, only Bands 1–2 may be fully printed
Electronics degradation rate	50% failure by Year 3–7 (estimate)	Faster degradation compresses the extraction window; slower gives more time
Workforce availability for curation	20–80 people (depends on competing demands)	Fewer people means slower extraction; may not complete lower-priority bands
Hardware condition at event	Unknown age distribution	If NZ’s hardware fleet is older than assumed, the window is shorter
Digital content accessibility	Varies by institution	NZ university library digital subscriptions (e.g., Elsevier, Springer, Wiley access via CONZUL consortium) are typically authenticated via internet-dependent licence servers; locally cached content may be DRM-locked. Crown Research Institute data may require institutional credentials. Access keys may be unavailable if key personnel are absent.
Quality of cached local content	Unknown	Much digital knowledge was accessed via internet subscriptions; what is actually cached locally vs. requiring internet access is unknown until surveyed
Paper stock condition	Depends on storage	Paper stocks may have degraded if stored in humid or uncontrolled conditions
Manual printing capability timeline	Year 2–5 for meaningful output (Doc #29)	If domestic paper and ink production is delayed, the gap between laser printing and manual printing widens

---

9. CROSS-REFERENCES

Document	Relationship
Doc #5 (Printing Supply Requisition)	Supply framework for toner, paper, and printer parts. Establishes the national printing authority that allocates resources to this programme.
Doc #8 (National Asset and Skills Census)	Establishes actual hardware inventory, toner stocks, and available workforce. Essential input for refining the estimates in this document.
Doc #156 (Skills Census)	Detailed skills census methodology — complements Doc #8 for workforce allocation to the extraction programme.
Doc #29 (National Printing Plan)	Paper and ink production infrastructure. Enables post-toner printing of extracted content.
Doc #30 (Print Optimization)	Format specifications for maximising output from finite printing supplies.
Doc #31 (Manual Printing Methods)	Operational guide for letterpress, screen printing, and stencil duplication — the methods used after toner depletion.
Doc #65 (Hydroelectric Maintenance)	Grid reliability determines how long digital storage and printers remain powered.
Doc #74 (Pastoral Farming)	Agricultural guidance that must be available in print for the farming sector.
Doc #116 (Pharmaceutical Rationing)	Medical content that depends on printed formulary access.
Doc #117 (Surgical Consumable Conservation)	Surgical references are Band 1 priority for extraction.
Doc #127 (NZ Telecommunications Maintenance)	Network availability affects access to distributed digital content across institutions.
Doc #129 (AI Inference Facility)	The AI facility can assist with content curation, prioritisation, and formatting of digital content for extraction.
Doc #133 (Local Network Architecture)	Local network infrastructure enables inter-institutional data transfer for content extraction.
Doc #130 (Device Life Extension)	Strategies for extending the operational life of NZ’s electronics — directly extends the extraction window.
Doc #139 (Celestial Navigation)	Navigation tables are Band 4 priority for extraction.
Doc #157 (Accelerated Trade Training)	Training curricula are Band 5 priority for extraction.
Doc #168 (Master Index)	This extraction schedule feeds into the master index of all printed knowledge holdings.

---

FOOTNOTES

---

1. Hard drive and SSD reliability data is well-documented by multiple sources. Backblaze, a cloud storage company, publishes annual drive failure statistics from its fleet of 200,000+ drives; as of 2024, annualised failure rates for drives in their 3rd–5th year of service ranged from 1–5% depending on model, with cumulative failure rates reaching 10–25% by year 5. Source: Backblaze Hard Drive Stats, https://www.backblaze.com/cloud-storage/resources/hard-dr... — SSD endurance depends on NAND type and write amplification; consumer SSDs typically carry 5-year warranties with defined write endurance (150–600 TBW for typical consumer drives). Stored unpowered, NAND flash cells lose charge over time — JEDEC standards specify data retention of 1 year at 30 degrees C for consumer SSDs and 3 months at 40 degrees C. Source: JEDEC JESD218 specification. Server component reliability data is available from manufacturer specifications and data centre operations literature.↩︎

2. The estimate that the majority of pre-war digital storage becomes unreadable by Year 10–15 is an inference from the component-level failure rates in footnote 1, applied to a mixed-age fleet without replacement parts. This is an estimate, not a measured fact. A well-maintained, climate-controlled archive of the best hardware could last longer; poorly maintained hardware in humid or dusty environments will fail sooner. The uncertainty range is large.↩︎

3. Recovery Library page count estimate from Doc #5, based on the 172-document catalogue at an average of 15–20 pages per document. Some documents (Doc #135, Computer Construction) are substantially longer; many are shorter.↩︎

4. M-DISC (Millenniata) optical media uses an inorganic recording layer claimed to last 1,000+ years under archival conditions. Standard DVD-R and BD-R media use organic dyes with estimated lifespans of 10–100+ years depending on dye type and storage conditions. Source: Library of Congress digital preservation research; M-DISC product specifications. For hard drives stored unpowered, the JEDEC specification for enterprise SSDs is 3 months at 40 degrees C — consumer drives at lower temperatures may retain data for 1–5+ years, but this is not guaranteed. Spinning drives stored unpowered may develop bearing issues or lubricant migration that prevents spin-up after extended storage.↩︎

5. NZ public library holdings: the National Library of New Zealand reports that NZ public libraries collectively hold approximately 8–12 million items across all formats (including AV materials, periodicals, and non-book items). Exact figures for monographs (books) are not readily separable from this total. Source: National Library of New Zealand, Public Libraries of New Zealand statistics. https://natlib.govt.nz/↩︎

6. NZ library and archives workforce: Stats NZ Household Labour Force Survey classifies librarians and related information professionals. The figure of 5,000–8,000 is an estimate based on published workforce data for the GLAM (galleries, libraries, archives, museums) sector. Exact figures require verification with Stats NZ.↩︎

7. NZ computer ownership: Stats NZ Household Use of ICT survey (most recent edition) reports that approximately 90% of NZ households have at least one computer (desktop, laptop, or tablet). With approximately 1.9 million households and some having multiple devices, an estimate of 2–3 million computers is reasonable. Business and institutional devices add significantly to this total. Source: Stats NZ, Household Use of Information and Communication Technology. https://www.stats.govt.nz/↩︎

8. Hard drive and SSD reliability data is well-documented by multiple sources. Backblaze, a cloud storage company, publishes annual drive failure statistics from its fleet of 200,000+ drives; as of 2024, annualised failure rates for drives in their 3rd–5th year of service ranged from 1–5% depending on model, with cumulative failure rates reaching 10–25% by year 5. Source: Backblaze Hard Drive Stats, https://www.backblaze.com/cloud-storage/resources/hard-dr... — SSD endurance depends on NAND type and write amplification; consumer SSDs typically carry 5-year warranties with defined write endurance (150–600 TBW for typical consumer drives). Stored unpowered, NAND flash cells lose charge over time — JEDEC standards specify data retention of 1 year at 30 degrees C for consumer SSDs and 3 months at 40 degrees C. Source: JEDEC JESD218 specification. Server component reliability data is available from manufacturer specifications and data centre operations literature.↩︎

9. Hard drive and SSD reliability data is well-documented by multiple sources. Backblaze, a cloud storage company, publishes annual drive failure statistics from its fleet of 200,000+ drives; as of 2024, annualised failure rates for drives in their 3rd–5th year of service ranged from 1–5% depending on model, with cumulative failure rates reaching 10–25% by year 5. Source: Backblaze Hard Drive Stats, https://www.backblaze.com/cloud-storage/resources/hard-dr... — SSD endurance depends on NAND type and write amplification; consumer SSDs typically carry 5-year warranties with defined write endurance (150–600 TBW for typical consumer drives). Stored unpowered, NAND flash cells lose charge over time — JEDEC standards specify data retention of 1 year at 30 degrees C for consumer SSDs and 3 months at 40 degrees C. Source: JEDEC JESD218 specification. Server component reliability data is available from manufacturer specifications and data centre operations literature.↩︎

10. Paper longevity: properly stored acid-free paper can last several hundred years. The Library of Congress and other major archives maintain documents on high-quality paper dating to the 15th century and earlier. Standard ISO 9706 specifies requirements for “permanent paper” suitable for archival use. Modern office paper that meets this standard is widely available. Source: Library of Congress preservation resources; ISO 9706:1994.↩︎

11. Two-column layout readability: research in typography and document design consistently shows that line lengths of 45–75 characters per line are optimal for reading speed and comprehension. A two-column A4 layout naturally produces line lengths in this range, while a single-column A4 layout at small font sizes produces lines of 80–100+ characters, which are harder to track visually. Source: Bringhurst, R. (2004), “The Elements of Typographic Style,” Hartley & Marks; general typography and readability research.↩︎

12. Toner consumption for graphics vs. text: standard laser printer toner coverage for text pages is approximately 5% of the page area. A full-page greyscale image with moderate tonal range may cover 30–80% of the page area, consuming roughly 5–15x the toner of a text page. The exact ratio depends on image density, greyscale distribution, and printer toner-save settings. Based on standard printer manufacturer specifications for page yield at 5% coverage (e.g., HP, Canon, and Xerox cartridge yield ratings).↩︎

13. The New Zealand Formulary (nzf.org.nz) is the national medicines reference, providing dosing, interaction, and prescribing information for all medicines available in NZ. It is primarily a digital resource, updated regularly online. Some printed editions have been produced historically (derived from the British National Formulary), but the current NZ-specific formulary is digital-first. Many NZ clinicians access it via smartphone or computer at the point of care. Source: NZ Formulary, https://nzf.org.nz/↩︎

14. The New Zealand Formulary (nzf.org.nz) is the national medicines reference, providing dosing, interaction, and prescribing information for all medicines available in NZ. It is primarily a digital resource, updated regularly online. Some printed editions have been produced historically (derived from the British National Formulary), but the current NZ-specific formulary is digital-first. Many NZ clinicians access it via smartphone or computer at the point of care. Source: NZ Formulary, https://nzf.org.nz/↩︎

15. Machinery’s Handbook (Industrial Press) is the standard reference for mechanical engineering workshop practice. Print editions exist and are still sold, but many NZ workshops have transitioned to the digital edition or use online equivalents. The 31st edition (2020) runs to approximately 2,800 pages. Source: Industrial Press, Machinery’s Handbook.↩︎

16. LINZ Topo50 map series: Land Information New Zealand produces the national topographic map series at 1:50,000 scale, covering all of NZ in approximately 450–500 map sheets. These are available digitally (free download) and in print (on demand or from selected retailers). Many NZ outdoor recreation retailers stock popular sheets, but a complete set of all sheets is not commonly held outside LINZ and specialised collections. Source: LINZ, https://www.linz.govt.nz/products-services/maps/↩︎

17. LINZ Topo50 map series: Land Information New Zealand produces the national topographic map series at 1:50,000 scale, covering all of NZ in approximately 450–500 map sheets. These are available digitally (free download) and in print (on demand or from selected retailers). Many NZ outdoor recreation retailers stock popular sheets, but a complete set of all sheets is not commonly held outside LINZ and specialised collections. Source: LINZ, https://www.linz.govt.nz/products-services/maps/↩︎

18. Te reo Maori digital resources: the revitalisation of te reo Maori since the 1980s has produced a substantial body of digital educational material — including resources from Te Taura Whiri i te Reo Maori (Maori Language Commission), Maori Television (online content), university language programmes, and community digital resources. The proportion of these resources that exists only in digital form (as opposed to also being available in print) is unknown but is believed to be significant for more recent materials. Source: Te Taura Whiri i te Reo Maori, https://www.tetaurawhiri.govt.nz/↩︎

19. NZ public library holdings: the National Library of New Zealand reports that NZ public libraries collectively hold approximately 8–12 million items across all formats (including AV materials, periodicals, and non-book items). Exact figures for monographs (books) are not readily separable from this total. Source: National Library of New Zealand, Public Libraries of New Zealand statistics. https://natlib.govt.nz/↩︎

20. NZ library shelving capacity: NZ has approximately 330 public library service points (National Library of New Zealand statistics). A medium-sized public library branch typically has 500–2,000 linear metres of shelving. University libraries (8 universities plus polytechnics) add substantial capacity. The figure of 200+ km total linear shelving nationally is a rough estimate; actual figures would be established by the skills census (Doc #8/Doc #156). Source: National Library of New Zealand, Public Libraries of New Zealand statistical reports.↩︎