Blockchain Infrastructure Demystified — What Cold Chain Operators Actually Need to Understand

Ask ten cold chain operators what blockchain does, and most will describe something like a tamper-proof digital lock on their temperature data — something they own, control, and run on their own systems. The mental model goes like this: I record my temperature data, blockchain seals it, done. Like writing a letter and putting it in a locked envelope that nobody can open.

This understanding is wrong in a way that matters for your budget.

Blockchain is not a feature you switch on inside your existing monitoring system. It is not a local encryption tool. It is not free after setup. And in most implementations, it requires ongoing participation in an external network — with all the costs, dependencies, and infrastructure questions that implies.

If you are evaluating monitoring platforms that market “blockchain-enabled” as a feature, or considering blockchain for export traceability, you need to understand what you are actually buying. Otherwise you are making investment decisions about infrastructure you do not understand — and that is exactly where vendors take advantage.

This article provides a plain-language explanation of how blockchain infrastructure actually works, the different deployment models available, what they cost, and what questions to ask before committing. No jargon where we can avoid it. Analogies where they help.

We explored where blockchain creates genuine value in cold chain operations in the second editorial of this series, and why IoT data quality must come first in the opening article. This piece addresses the question those articles deliberately left for a dedicated technical guide: how does blockchain actually work under the hood, and what does it cost to run?

Think of it this way. Blockchain is not the envelope — it is the postal system. You still write the letter (your sensor data), you still seal it (data integrity), but blockchain is the network that ensures every party gets an identical copy that nobody can alter after sending, and everyone can verify it was delivered intact. Like any postal system, it has infrastructure, operating costs, and dependencies you need to understand before you commit.

The Three Infrastructure Models — And Why They Matter

When a vendor says “our platform uses blockchain,” they could mean one of three very different things. Understanding which model they are actually offering is the single most important question you can ask before signing anything.

Model 1: Public Blockchain — The Global Network

The analogy: Using the national postal service. Anyone can send, anyone can receive, everyone pays postage per letter. You do not own the post offices.

A public blockchain is a global, open network that anyone can join. Ethereum is the most well-known example. Your cold chain data gets recorded as transactions on this shared global ledger, maintained by thousands of independent computers (called nodes) around the world through consensus mechanisms — essentially, mathematical processes that allow these computers to agree on which transactions are valid without any single authority making the decision.

You access a public blockchain through APIs or gateway services. You do not run a node yourself — you connect through a provider. And you pay per transaction through what Ethereum calls “gas fees,” which fluctuate based on network congestion. On Ethereum’s base layer, a single transaction can cost anything from fractions of a cent to several dollars. Layer-2 solutions — secondary networks built on top of Ethereum to improve speed and reduce costs — can bring per-transaction costs close to zero, but they add architectural complexity.

The strength of public blockchain is maximum transparency and independence. No single company controls the network, and your data persists even if your vendor disappears. The trade-offs include slower processing speeds, public visibility of transaction metadata (not necessarily your data content, but the fact that transactions occurred), and the energy consumption concerns that have dogged public blockchains for years — though Ethereum’s 2022 shift from proof-of-work to proof-of-stake reduced its energy footprint by roughly 99%.

Cold chain reality check: Very few cold chain implementations use public blockchains directly. The parties in a cold supply chain are known entities — your transport company, your cold store, your export agent, the retailer. You do not need the whole world to verify your temperature records. Public blockchain is overkill for most cold chain use cases.

Model 2: Consortium Blockchain — The Members-Only Network

The analogy: A private courier service run jointly by a group of companies who all need to exchange verified documents. They share the infrastructure costs and agree on the rules.

A consortium blockchain is a network operated by a defined group of participants. Only approved members can read or write data. Hyperledger Fabric, an open-source framework hosted by the Linux Foundation, is the dominant platform for enterprise consortium blockchains. It powers some of the most cited supply chain implementations globally, including the IBM Food Trust network used by Walmart.

Each consortium member — or their nominated cloud provider — runs a node. A governing body sets the rules for participation, data access, and network upgrades. You access the network through your node or via API. Membership usually requires agreement to governance terms and technical standards.

The cost structure is fundamentally different from public blockchain. There are no volatile gas fees. Instead, costs come from infrastructure (running or hosting nodes), consortium membership fees, and governance participation. These costs are typically shared among members.

Performance is dramatically better than public blockchain. Research published in the International Journal of Network Management demonstrated that Hyperledger Fabric can be scaled to process 20,000 transactions per second under optimised conditions — orders of magnitude faster than public blockchain networks. Practically, most enterprise deployments achieve hundreds to several thousand transactions per second, which is more than adequate for cold chain monitoring.

Privacy is built in: data is visible only to permitted parties. And the consortium collectively controls the rules — there is no external authority dictating terms.

The downsides are structural. Someone has to organise the consortium, establish governance, and coordinate members. The network is less “trustless” than public blockchain — you are trusting the consortium members, not a mathematical proof. And the network’s value depends entirely on participation: if key members leave, the value diminishes.

Cold chain reality: This is where most serious supply chain blockchain deployments sit. Walmart’s food safety blockchain reduced the time needed to trace mango provenance from seven days to 2.2 seconds using Hyperledger Fabric — a result that fundamentally changed the business case for blockchain in food safety. For South Africa’s cold chain, a consortium of exporters, the PPECB, cold stores, and shipping lines would be the logical structure. But it requires industry coordination to establish.

Model 3: Blockchain-as-a-Service / BaaS — The Cloud-Hosted Option

The analogy: Renting a post office box from a managed service provider instead of building your own postal infrastructure.

BaaS is exactly what it sounds like: cloud providers host and manage blockchain nodes for you. You subscribe, configure, and deploy without managing the underlying infrastructure. Amazon Web Services (AWS) Managed Blockchain, IBM Blockchain Platform, and Oracle Blockchain Platform are the major enterprise offerings.

You access BaaS through the provider’s cloud dashboard and APIs, just like any other cloud service. Pricing is subscription-based plus usage fees — compute, storage, and data transfer. AWS Managed Blockchain, for example, charges per-hour rates for network membership and peer nodes, plus storage and data transfer costs. A Standard Edition membership on AWS runs approximately $0.30 per hour, with peer node costs from $0.034 per hour for the smallest instance — costs that accumulate quickly in production environments.

The BaaS market is growing rapidly. SkyQuest Technology estimated it at USD 3.67 billion in 2024, projecting growth to USD 51.8 billion by 2032 — a compound annual growth rate of 39.2%. This growth reflects enterprise demand for blockchain capability without the complexity of building infrastructure from scratch.

BaaS offers the fastest deployment path, the lowest technical barrier, enterprise support, and integration with existing cloud services. For South African cold chain operators who want blockchain capability without building infrastructure, BaaS is the most realistic entry point.

But there is a critical caveat. If a single vendor hosts all the nodes on their own cloud instance, you have essentially recreated a centralised database with extra steps. The “trust” value of blockchain — the reason you are paying for it rather than just using a regular database — only works when multiple independent parties operate nodes. A vendor running blockchain on their own servers and calling it “blockchain-enabled” is a marketing claim, not a trust architecture.

And BaaS creates its own dependency. You are adopting another SaaS relationship with ongoing costs and vendor lock-in — exactly the kind of dependency that blockchain was philosophically designed to eliminate.

The Key Insight for Operators

When a monitoring vendor says “blockchain-enabled,” ask which model they use. If it is a single vendor running all the nodes on their own cloud instance, that is a database with blockchain branding — not a trust infrastructure. The postal system analogy holds: if one company owns every post office, every sorting facility, and every delivery truck, you do not have an independent postal system. You have a courier company with a fancy name.

The Cost Reality — What Blockchain Actually Costs to Implement

The blockchain sales pitch often glosses over ongoing costs. Operators need to understand cost structures before they sign.

Infrastructure Costs

The costs vary dramatically by model:

• Public blockchain: Gas fees per transaction. For cold chain monitoring recording 100 temperature events per vehicle per day across a 20-vehicle fleet, that is roughly 2,000 transactions per day. On Ethereum Layer-2 solutions, this could cost fractions of a cent per transaction — very low. On Ethereum’s base layer during periods of network congestion, it could be several dollars per transaction — prohibitively expensive at scale.
• Consortium blockchain: Node hosting plus compute, storage, and network costs. If using cloud hosting through AWS or similar providers, budget R5,000–R25,000 per month per node depending on transaction volume and data retention requirements.
• BaaS: Subscription fees on top of infrastructure. Budget R10,000–R50,000 per month for a meaningful enterprise deployment, though costs vary dramatically by provider, scale, and configuration.

Integration Costs

Connecting your existing monitoring platform to a blockchain network requires API development or middleware. This is often the largest cost and the one most frequently underestimated. Budget R100,000–R500,000 or more for initial integration depending on complexity. Integration with existing ERP and warehouse management systems adds further cost layers.

Ongoing Operational Costs

Data storage grows continuously — temperature readings accumulate, and blockchain’s immutable nature means nothing gets deleted. Add network participation costs (membership fees for consortiums), technical support and maintenance, staff training, and regular audits and compliance verification.

Consortium Participation Costs

If you join a consortium model, factor in governance participation (time and resources), legal agreements and compliance reviews, and technology alignment costs when the consortium evolves its technical standards.

The Honest Comparison

For a medium-sized South African cold chain operator running 20 vehicles and 2 cold stores, the picture looks roughly like this:

Cost Component	Monthly Estimate (ZAR)
Cloud-based IoT monitoring platform (without blockchain)	R15,000 – R30,000
Adding blockchain through BaaS	R25,000 – R75,000 additional
Total with blockchain	R40,000 – R105,000

The blockchain layer roughly doubles to triples your monitoring costs, once infrastructure and integration expenses are amortised. Is the ROI there? Only if the blockchain layer creates value you cannot achieve otherwise — typically export traceability verification, multi-party chain-of-custody records, or retailer requirements that specifically demand blockchain-verified documentation.

For purely internal temperature monitoring where you are the only party that needs to trust the data, a well-configured cloud IoT platform with proper access controls and audit logs delivers the same practical outcome at a fraction of the cost.

The Dependency Question — What Your Operation Relies On

Every technology creates dependencies. With blockchain, the dependencies are different from what operators are used to, and they are rarely discussed during the sales process.

Network Dependency

On a public blockchain, the network must remain operational — Ethereum is not going anywhere, but smaller chains might. In a consortium, all members must continue participating. If key members leave, the network’s verification value diminishes. With BaaS, you are dependent on the cloud provider’s continued service and pricing — and pricing can change.

This is not theoretical. Microsoft retired its Azure Blockchain Service on 10 September 2021, giving customers just four months’ notice to migrate their ledger data to alternative offerings. Customers who had built compliance infrastructure on Azure Blockchain were forced to migrate to ConsenSys Quorum or rebuild entirely. If that provider was the foundation of your export traceability records, those four months would have been extremely stressful.

Internet Connectivity

Blockchain transactions require connectivity at the point of recording — or buffered for later submission. In South Africa, this creates immediate practical questions. What happens when your gateway loses connectivity during load shedding for four hours? Does the system buffer data locally and submit later (store-and-forward capability), or do you have gaps in your blockchain record? For cross-border SADC routes, satellite connectivity for remote segments adds cost but may be necessary for continuous blockchain recording.

Any blockchain implementation for South African cold chain operations must account for the connectivity reality. If the system cannot handle offline periods gracefully, it is not fit for purpose in this market.

Data Sovereignty

Where are the blockchain nodes physically hosted? South Africa’s Protection of Personal Information Act (POPIA) regulates cross-border transfers of personal information under Section 72, requiring that recipient countries provide adequate levels of data protection. While POPIA does not prohibit cross-border data flows outright, it creates conditions that must be satisfied.

If you are using AWS or similar cloud providers, are nodes hosted in South Africa, Europe, or US data centres? AWS operates a Cape Town region (af-south-1), which supports local data residency — but not all BaaS configurations default to local hosting. For pharmaceutical cold chain operations governed by SAHPRA’s Good Distribution Practice requirements, data sovereignty may be a compliance factor that needs explicit verification.

Consortium models can specify node locations as part of their governance rules. BaaS models depend on which cloud regions the provider offers — and you need to confirm this before signing, not discover it during an audit.

Vendor Continuity

The Microsoft Azure Blockchain retirement is a cautionary tale worth repeating. Ask any prospective blockchain vendor: what happens to my verified records if you cease operations, exit the blockchain market, or fundamentally change your pricing? If the answer is not clear and contractually documented, you are building compliance infrastructure on uncertain foundations.

Data portability matters too. Can you export your blockchain-verified records in a usable, standard format — such as EPCIS 2.0 — if you need to migrate? Or are you locked into a proprietary data structure that dies with the platform?

Smart Contracts — The Automation Layer Explained Simply

Operators often hear “smart contracts” and imagine artificial intelligence making decisions about their cold chain. The reality is far more mundane — and more useful for it.

• What operators think smart contracts are: AI that makes decisions about their cold chain.
• What smart contracts actually are: Pre-programmed rules that execute automatically when conditions are met.
• The analogy: A vending machine. You put in the correct payment (condition met), and it dispenses the product (action executes). No human decision-making involved. It follows the rules it was programmed with — nothing more, nothing less.

Here is what cold chain smart contracts look like in plain language:

• IF temperature reading exceeds +8°C for more than 30 minutes → THEN flag shipment as “excursion recorded” and notify all parties in the chain
• IF shipment arrives at destination AND all temperature readings are within range → THEN update status to “cold chain verified” and release payment
• IF handoff from Transport Company A to Cold Store B is recorded → THEN transfer chain-of-custody responsibility on the ledger

Smart contracts require reliable input data (back to the garbage-in-garbage-out problem), clear and unambiguous rules written as code rather than subjective judgment, parties who agree to the rules in advance, and a blockchain network to execute on.

They do not make judgment calls about edge cases, replace human decision-making for complex situations, work better than the data feeding them, or cost nothing to deploy and maintain.

For South African cold chain operations, smart contracts have genuine potential for export documentation automation — if temperature records are clean throughout the chain, auto-generate the compliance certificate — and for payment triggering, where verified cold chain compliance triggers payment release. But these are Phase 2–3 capabilities, not a starting point. Get your data quality and basic blockchain infrastructure right first.

Twelve Questions to Ask Before You Invest

Before signing anything with “blockchain” in the proposal, work through this evaluation framework. These questions are designed to separate genuine blockchain value from marketing theatre.

Infrastructure Questions

1. Which blockchain model does your platform use? Public, consortium, or BaaS — if the vendor cannot answer clearly, walk away.

2. Who operates the nodes? If the vendor alone operates all nodes, it is a centralised database with blockchain branding. The trust value of blockchain requires multiple independent node operators.

3. Where are the nodes physically hosted? This determines data sovereignty for POPIA and SAHPRA GDP compliance. Get this in writing.

4. What happens to my data if I leave the platform or if you cease operations? Demand a clear, contractually documented answer. Remember Microsoft Azure Blockchain.

Cost Questions

5. What are the total monthly costs? Break it down: subscription fees, transaction fees, storage, compute, and data transfer. Get the complete picture, not just the headline subscription number.

6. How do costs scale as my data volume grows? A fleet of 100 sensors recording every 5 minutes generates significant data accumulation over months and years. Understand the cost curve, not just the entry price.

7. What are the integration costs to connect with my existing monitoring, WMS, and TMS systems? This is often the largest single expense and the one most vendors minimise during sales conversations.

8. Are there consortium membership or governance fees? For consortium models, these ongoing costs are separate from infrastructure and can be substantial.

Technical Questions

9. Does the platform support EPCIS 2.0 data format? If not, you are locking into a proprietary data structure on top of a blockchain platform — double lock-in. EPCIS 2.0 is the emerging global standard for supply chain event data, and adopting it now protects your future interoperability.

10. How does the system handle connectivity gaps? Store-and-forward capability during load shedding or cellular dead zones is non-negotiable for South African operations. Test this claim, do not just accept it.

11. Can I export my blockchain-verified records in open formats for independent verification? If the answer is no, your “independent verification” depends entirely on one vendor’s continued existence.

12. What consensus mechanism is used, and what are the performance implications? In practical terms: how many transactions per second does the network handle, and what latency should you expect? For cold chain monitoring at scale, you need this to work reliably, not just in a demo environment.

The Bottom Line Question

“What specific trust problem does this solve that my current cloud monitoring platform with proper access controls and audit logs does not?”

If the answer is primarily about multi-party verification in supply chains with three or more independent entities — blockchain may be worth the investment. If the answer is about making your own internal data “more secure” — you probably do not need blockchain. You need better access controls.

Where This Leaves South African Operators

Blockchain infrastructure is not magic and it is not free. It is a specific type of distributed database architecture that creates value when multiple independent parties need to share verified records without trusting each other’s internal systems.

For South African cold chain operators, the practical reality comes down to this:

Most internal monitoring needs are better served by well-configured cloud IoT platforms with proper security and audit trails. The cost differential is significant, and the trust problem blockchain solves — verifying data between parties who do not trust each other’s systems — does not apply to your own internal operations.

Blockchain’s value emerges in multi-party supply chains. Export corridors where EU traceability requirements are tightening under frameworks like the EUDR, SADC cross-border verification that is still developing, and retailer transparency demands that increasingly specify blockchain-verified records — these are the use cases where the infrastructure investment starts to make economic sense.

The most realistic entry point for most operators is BaaS, but go in with your eyes open. You are adopting another SaaS dependency with ongoing costs — not installing a one-time security feature. Budget accordingly, and make sure your contract addresses data portability and vendor continuity.

The strongest value proposition sits with consortium blockchains, but they require industry coordination to establish. This is where bodies like the CGCSA, PPECB, and GCCA South Africa could play a catalytic role — establishing shared blockchain infrastructure that individual operators could join, rather than each company trying to build or subscribe independently.

The operators who will make good blockchain investment decisions are those who first understand what they are buying. Blockchain is not the envelope — it is the postal system. Make sure you need a postal system before you pay for one. And if you do need one, make sure you understand who runs the post offices, what the postage costs, and what happens to your letters if the postal service shuts down.

For a comprehensive overview of the regulatory requirements that blockchain implementation must satisfy, see our South Africa Cold Chain Compliance Matrix. And browse our Technology directory for monitoring and blockchain platform providers serving the South African market.

Sources & References

About These Sources

This article draws on authoritative sources including blockchain framework documentation (Hyperledger, Ethereum), cloud provider pricing data (AWS), peer-reviewed academic research, industry market reports (SkyQuest, Mordor Intelligence, Gartner), and South African regulatory frameworks (POPIA). All sources were verified as of February 2025 and represent the most current publicly available information on blockchain infrastructure for supply chain applications.

Citation Methodology

Direct data points in the article reference these sources. Where cost estimates are contextualised for South African operations (Rand values, load shedding impacts, POPIA considerations), the article draws on operational experience and industry-specific analysis. Readers seeking additional detail on any cited statistic can access the source material directly through the URLs provided.

Currency Note

BaaS market projections, enterprise blockchain spending estimates, and cloud provider pricing reflect published data as of early 2025. Cloud service pricing changes frequently — verify current rates directly with providers before making investment decisions. Rand-denominated cost estimates are indicative ranges based on current exchange rates and typical deployment configurations, not binding quotations.

Blockchain Infrastructure & Frameworks

• Hyperledger Fabric Documentation — Introduction and Architecture Overview — Open-source enterprise blockchain framework hosted by the Linux Foundation, providing technical specifications for permissioned distributed ledger networks.
• Amazon Web Services — Amazon Managed Blockchain Pricing — Current pricing structure for AWS’s blockchain-as-a-service offering, including Hyperledger Fabric and Ethereum node costs.
• Amazon Web Services — Managed Blockchain for Hyperledger Fabric Pricing — Detailed pricing examples for Starter and Standard Edition Hyperledger Fabric networks on AWS.
• Ethereum Foundation — Proof-of-Stake and Layer-2 Documentation — Technical documentation on Ethereum’s consensus mechanism transition and scaling solutions.
• InfoQ — Microsoft Retires Azure Blockchain (June 2021) — Coverage of Microsoft’s decision to retire Azure Blockchain Service, including the migration timeline and industry implications.

Market Data & Industry Analysis

• SkyQuest Technology — Blockchain as a Service Market Report — BaaS market valuation at USD 3.67 billion (2024) with projections to USD 51.8 billion by 2032 at 39.2% CAGR.
• Mordor Intelligence — Blockchain-as-a-Service Market Size, Share & Trend Report (2025–2030) — Enterprise BaaS adoption analysis including deployment model breakdown and vertical market segmentation.
• Gartner — Identifies the Four Phases of the Blockchain Spectrum — Forecast of blockchain generating $3.1 trillion in new business value by 2030, with analysis of enterprise adoption phases.

Supply Chain Implementation Case Studies

• Linux Foundation Decentralised Trust — Walmart Hyperledger Fabric Case Study — Detailed case study of Walmart’s blockchain-based food traceability system reducing mango provenance trace time from seven days to 2.2 seconds.
• Walmart Global Tech — Blockchain in the Food Supply Chain — First-hand account of Walmart’s blockchain implementation for food safety, including pork traceability in China and mango tracing in the US.
• Gorenflo et al. — FastFabric: Scaling Hyperledger Fabric to 20,000 Transactions per Second (International Journal of Network Management, 2020) — Peer-reviewed research demonstrating Hyperledger Fabric performance optimisation for enterprise-scale throughput.

South African Regulatory Context

• Protection of Personal Information Act (POPIA) — Full Act Text — South Africa’s primary data protection legislation, including Section 72 governing cross-border transfers of personal information.
• InCountry — South Africa’s Data Sovereignty Laws and Regulations — Analysis of POPIA, the National Policy on Data and Cloud (2024), and their implications for cloud-hosted data services.
• Amazon Web Services — South Africa Data Privacy Compliance — AWS documentation on POPIA compliance measures and the Cape Town (af-south-1) data centre region for local data residency.

BaaS Provider & Platform Comparisons

• IT Pro — Microsoft to Shut Down Azure Blockchain Service (May 2021) — Detailed reporting on Azure Blockchain Service retirement after only two years, including customer impact and migration challenges.
• Amazon Web Services — Amazon Managed Blockchain Features — Technical specifications for AWS’s blockchain infrastructure offering, including Hyperledger Fabric and Ethereum support.

Academic & Critical Analysis

• ScienceDirect — In-Depth Investigation of Performance Characteristics of Hyperledger Fabric (Computers & Industrial Engineering, 2022) — Comprehensive performance analysis of Hyperledger Fabric for enterprise applications, covering throughput, latency, and scalability metrics.
• Journal of the British Blockchain Association — Food Traceability on Blockchain: Walmart’s Pork and Mango Pilots with IBM — Peer-reviewed analysis of Walmart’s blockchain pilot programs, challenges, and implications for food supply chain traceability.

Related Resources

• Why Your Cold Chain IoT Sensors Can’t Talk to Each Other — And What’s Coming to Fix It — The IoT interoperability problem and what global standards like EPCIS 2.0 mean for SA operators
• Blockchain in Cold Chain: What South African Operators Actually Need to Know in 2025 — Where blockchain creates real value versus marketing hype in cold chain applications
• South Africa’s Cold Chain Compliance Matrix — Every regulation, certification, and requirement mapped by operator type
• Cold Chain Glossary — Technical terminology and definitions for the cold chain industry
• Temperature Monitoring & Technology Directory — Browse monitoring, blockchain, and IoT platform providers serving South African cold chain operators

About ColdChainSA

ColdChainSA.com is South Africa’s dedicated cold chain industry directory and knowledge platform. All editorial content is independent and evidence-based. Directory listings are presented on equal terms regardless of commercial relationships.